kayg/calternal

Watch

Fork 0

WIP: calendar v2 slice 2 optimized in-process parser gates #805

Closed

kayg wants to merge 0 commits from feat/calendar-v2-s2-parser-contract into dev

kayg commented

2026-10-31 18:24:43 +00:00

Owner

Round 2 implementation status

Head: 948117ccc9dbcfa02e044dc5efc6484a32f414e1. Never merge. No repository settings changed. Ready for the independent final review.

Item	Status	Evidence
P1 bare-CR reproduction	PASS	Before: `x\r` × 2,097,152 exits -6 with allocation failure under 256 MiB AS / 8 MiB stack. After: unresolved/line-limit. Intermediate 1 MiB and 4 MiB CR, LF, CRLF and mixed cases safely refuse.
One normalized line view	PASS	One allocation-free iterator defines LF, CRLF and lone CR as breaks, including a final empty line. Every line/blank-run/length/syntax admission count, canonical LF materialization and walk line array derives from it. Parser ranges map back to original byte offsets before extraction. Text and offset-map growth charge the shared budget first.
Memory property / cap edges	PASS	72 seeded independent Linux child cases in crate tests, real 256 MiB AS / 8 MiB stack, 32 MiB VmHWM guard; limit setup must succeed and both accepted/refused cases must occur. Separately: 136 release probes, including 72 generated mixed-ending/blank-run/nesting cases, cap edges and density searches; peak 30,052 KiB, below unchanged 32 MiB RSS guard.
P2 four metadata disagreements	PASS	Four independently authored shared vectors pass both adapters with the editor's existing decisions: #tag is extracted after word, word, word, and inside word#tag. Both round-1 fix vectors are unchanged.
Original offsets / line semantics	PASS	Shared Unicode/mixed-ending/frontmatter/task/child-link vectors check exact spans and source slices in both adapters. Lone CR always breaks a line, so the old title-trim-0xd vector intentionally changes to three lines with an empty first-event title.
Native gates	PASS	25 crate tests; original 70 Rust line vectors; 12 Composer create tests; fmt; clippy -D warnings.
Editor gates	PASS	Locked install, typecheck, 207 tests across 3 contract files.
CI memory regression	PASS on VM	Executed the exact workflow LF/bare-CR/mixed shell probes under ulimit, asserting unresolved reasons. Native crate property also runs in the native CI job. No conditional contract skip or continue-on-error.
Speed budget	PASS	4/32/256 KiB p95 growth +7.31% / -2.14% / -9.69%; added p95 +0.003538 / -0.007543 / -0.315301 ms. Paired 100k wall 5.457216405s direct / 5.526577250s head (+1.27%).
Hosted checks / final review	NOT CLAIMED	Forgejo reports no commit status contexts; VM gates are separate evidence. Independent final review remains pending.

All source/admission/event/result caps remain unchanged. The first three speed runs failed and are retained in the plan evidence. Callgrind identified character-search/iterator overhead; the final implementation uses the already-locked memchr package, an inline shared iterator, a direct unmapped-range path, and removes the line-length array made redundant by normalization. No budget or fixture changed.

Same original direct baseline 9ce09e32, seed-805 byte identity for 100 variants, original two shard hashes, two lanes, warm 10k then paired 100k, nearest-rank quantiles. Zero unresolved, 925,696,000 bytes, 200,000 rows and 50,000 relationships. Release hashes verified on dev-01, control-host relay and perf-02:

577d893232a852172c514f0cab37b5d89c1a7509d1616575eb68730fb8869c65 parser_options
d95600e5c25ba09c2c50faa38631b170d6e47e8ecfb80d44e358d32f90f36229 in_process_containment

Exact dev-01 gates, in Nix, each heavy command through ~/bin/slot:

cargo test -j4 -p calternal-markdown-view --locked
cargo test -j4 --locked --manifest-path docs/plans/0067-vectors/rust/Cargo.toml
cargo fmt -p calternal-markdown-view --check
cargo clippy -j4 -p calternal-markdown-view --all-targets --locked -- -D warnings
cargo test -j4 -p calternal-tasks --locked create_tests::
cargo build -j4 --release -p calternal-markdown-view --examples --locked
bun install --frozen-lockfile
bun run --cwd packages/editor check
# From packages/editor:
bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts

Perf-02, under perf-lock:

~/bin/perf-lock python3 ~/calendar805-r2/check-markdown-memory.py ~/calendar805-r2/in_process_containment
~/bin/perf-lock python3 ~/calendar805-r2/compare-optimized.py ~/calendar805-r2

Limits: the 50 ms deadline is cooperative; the 64 MiB working counter remains a reservation model rather than a process-wide allocator. Memory probes relax only time to 30s to exercise full allocation and cleanup. wait4 RSS may conservatively include the parent's pre-exec high-water mark. No numeric worst-case time, unrelated allocation-pressure recovery, or consumer/publication integration is claimed.

Plan, decisions, failure table, exact results and retained failed runs. Runtime/contract file hashes match the tested VM files. Raw final logs: dev-01 ~/calendar805-r2-final.log, ~/calendar805-r2-ci-final.log; perf-02 ~/calendar805-r2-memory4.log, ~/calendar805-r2-perf4.log. Original failing probes: dev-01 ~/calendar805-r2-red.log, ~/calendar805-r2-red2.log.

Every meaningful step committed and pushed. No control-host builds/tests/installs; no merge.

## Round 2 implementation status Head: `948117ccc9dbcfa02e044dc5efc6484a32f414e1`. **Never merge.** No repository settings changed. Ready for the independent final review. | Item | Status | Evidence | | --- | --- | --- | | P1 bare-CR reproduction | PASS | Before: `x\r` × 2,097,152 exits -6 with allocation failure under 256 MiB AS / 8 MiB stack. After: unresolved/line-limit. Intermediate 1 MiB and 4 MiB CR, LF, CRLF and mixed cases safely refuse. | | One normalized line view | PASS | One allocation-free iterator defines LF, CRLF and lone CR as breaks, including a final empty line. Every line/blank-run/length/syntax admission count, canonical LF materialization and walk line array derives from it. Parser ranges map back to original byte offsets before extraction. Text and offset-map growth charge the shared budget first. | | Memory property / cap edges | PASS | 72 seeded independent Linux child cases in crate tests, real 256 MiB AS / 8 MiB stack, 32 MiB VmHWM guard; limit setup must succeed and both accepted/refused cases must occur. Separately: 136 release probes, including 72 generated mixed-ending/blank-run/nesting cases, cap edges and density searches; peak 30,052 KiB, below unchanged 32 MiB RSS guard. | | P2 four metadata disagreements | PASS | Four independently authored shared vectors pass both adapters with the editor's existing decisions: #tag is extracted after *word*, **word**, [word](p), and inside word[#tag](p). Both round-1 fix vectors are unchanged. | | Original offsets / line semantics | PASS | Shared Unicode/mixed-ending/frontmatter/task/child-link vectors check exact spans and source slices in both adapters. Lone CR always breaks a line, so the old title-trim-0xd vector intentionally changes to three lines with an empty first-event title. | | Native gates | PASS | 25 crate tests; original 70 Rust line vectors; 12 Composer create tests; fmt; clippy -D warnings. | | Editor gates | PASS | Locked install, typecheck, 207 tests across 3 contract files. | | CI memory regression | PASS on VM | Executed the exact workflow LF/bare-CR/mixed shell probes under ulimit, asserting unresolved reasons. Native crate property also runs in the native CI job. No conditional contract skip or continue-on-error. | | Speed budget | PASS | 4/32/256 KiB p95 growth +7.31% / -2.14% / -9.69%; added p95 +0.003538 / -0.007543 / -0.315301 ms. Paired 100k wall 5.457216405s direct / 5.526577250s head (+1.27%). | | Hosted checks / final review | NOT CLAIMED | Forgejo reports no commit status contexts; VM gates are separate evidence. Independent final review remains pending. | All source/admission/event/result caps remain unchanged. The first three speed runs failed and are retained in the plan evidence. Callgrind identified character-search/iterator overhead; the final implementation uses the already-locked memchr package, an inline shared iterator, a direct unmapped-range path, and removes the line-length array made redundant by normalization. No budget or fixture changed. Same original direct baseline `9ce09e32`, seed-805 byte identity for 100 variants, original two shard hashes, two lanes, warm 10k then paired 100k, nearest-rank quantiles. Zero unresolved, 925,696,000 bytes, 200,000 rows and 50,000 relationships. Release hashes verified on dev-01, control-host relay and perf-02: ```text 577d893232a852172c514f0cab37b5d89c1a7509d1616575eb68730fb8869c65 parser_options d95600e5c25ba09c2c50faa38631b170d6e47e8ecfb80d44e358d32f90f36229 in_process_containment ``` Exact dev-01 gates, in Nix, each heavy command through `~/bin/slot`: ```sh cargo test -j4 -p calternal-markdown-view --locked cargo test -j4 --locked --manifest-path docs/plans/0067-vectors/rust/Cargo.toml cargo fmt -p calternal-markdown-view --check cargo clippy -j4 -p calternal-markdown-view --all-targets --locked -- -D warnings cargo test -j4 -p calternal-tasks --locked create_tests:: cargo build -j4 --release -p calternal-markdown-view --examples --locked bun install --frozen-lockfile bun run --cwd packages/editor check # From packages/editor: bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts ``` Perf-02, under perf-lock: ```sh ~/bin/perf-lock python3 ~/calendar805-r2/check-markdown-memory.py ~/calendar805-r2/in_process_containment ~/bin/perf-lock python3 ~/calendar805-r2/compare-optimized.py ~/calendar805-r2 ``` Limits: the 50 ms deadline is cooperative; the 64 MiB working counter remains a reservation model rather than a process-wide allocator. Memory probes relax only time to 30s to exercise full allocation and cleanup. wait4 RSS may conservatively include the parent's pre-exec high-water mark. No numeric worst-case time, unrelated allocation-pressure recovery, or consumer/publication integration is claimed. [Plan, decisions, failure table, exact results and retained failed runs](https://git.kayg.org/kayg/calternal/src/branch/feat/calendar-v2-s2-parser-contract/docs/plans/0067-slice2/README.md). Runtime/contract file hashes match the tested VM files. Raw final logs: dev-01 `~/calendar805-r2-final.log`, `~/calendar805-r2-ci-final.log`; perf-02 `~/calendar805-r2-memory4.log`, `~/calendar805-r2-perf4.log`. Original failing probes: dev-01 `~/calendar805-r2-red.log`, `~/calendar805-r2-red2.log`. Every meaningful step committed and pushed. No control-host builds/tests/installs; no merge.

optimization-profile-evidence.txt

35 KiB

rules-red-evidence.txt

32 KiB

README.md

71 KiB

kayg added 1 commit

2026-10-31 18:24:43 +00:00

test(calendar): seed slice 2 with approved parser parity spike

CI / server (pull_request) Has been cancelled

Details

CI / rust (pull_request) Has been cancelled

Details

CI / web (pull_request) Has been cancelled

Details

fec18a59de

kayg changed title from ~~WIP: calendar v2 slice 2 parser contract gates~~ to WIP: calendar v2 slice 2 blocked on parser/Composer contract

2026-10-31 18:31:41 +00:00

kayg changed title from ~~WIP: calendar v2 slice 2 blocked on parser/Composer contract~~ to WIP: calendar v2 slice 2 scoped extraction contract

2026-10-31 18:41:09 +00:00

kayg changed title from ~~WIP: calendar v2 slice 2 scoped extraction contract~~ to WIP: calendar v2 slice 2 blocked on extraction spans

2026-10-31 18:55:07 +00:00

kayg changed title from ~~WIP: calendar v2 slice 2 blocked on extraction spans~~ to WIP: calendar v2 slice 2 blocked on native depth containment

2026-10-31 19:09:39 +00:00

kayg changed title from ~~WIP: calendar v2 slice 2 blocked on native depth containment~~ to WIP: calendar v2 slice 2 blocked on native node cap

2026-10-31 19:37:38 +00:00

kayg changed title from ~~WIP: calendar v2 slice 2 blocked on native node cap~~ to WIP: calendar v2 slice 2 blocked on parser/IPC budget

2026-10-31 19:56:05 +00:00

kayg changed title from ~~WIP: calendar v2 slice 2 blocked on parser/IPC budget~~ to WIP: calendar v2 slice 2 measuring in-process / hybrid options

2026-10-31 20:02:48 +00:00

kayg changed title from ~~WIP: calendar v2 slice 2 measuring in-process / hybrid options~~ to WIP: calendar v2 slice 2 blocked on A/C parser budgets

2026-10-31 20:10:26 +00:00

kayg commented

2026-10-31 20:16:04 +00:00

Author

Owner

Architecture decision: option A (in-process bounded extraction with catch_unwind), as directed by the orchestrator. Never merge.

Optimization plan: profile fixed 32/256 KiB seed-805 sources against direct baseline 9ce09e32, preserve O(source + events + rows + result) and all exclusive limits/authored spans. perf recording is denied by kernel perf_event_paranoid=3; use Nix Callgrind if no permitted perf event is available. Reduce measured hot costs, then re-run unchanged two-lane paired 10k/100k harness and nearest-rank quantiles on perf-02 under perf-lock with SHA-verified binaries. No worker/consumer migration or budget change.

Failure table remains the PR description table: source/depth/node/row/result/deadline refusal returns no payload; panic returns panic; native allocation/stack aborts are not contained. Tests: unchanged authored contracts, every exclusive counter/result boundary, seeded fuzz, depth and unwind recovery. If optimized larger buckets still fail, stop and report profile and numbers.

Architecture decision: option A (in-process bounded extraction with catch_unwind), as directed by the orchestrator. Never merge. Optimization plan: profile fixed 32/256 KiB seed-805 sources against direct baseline 9ce09e32, preserve O(source + events + rows + result) and all exclusive limits/authored spans. perf recording is denied by kernel perf_event_paranoid=3; use Nix Callgrind if no permitted perf event is available. Reduce measured hot costs, then re-run unchanged two-lane paired 10k/100k harness and nearest-rank quantiles on perf-02 under perf-lock with SHA-verified binaries. No worker/consumer migration or budget change. Failure table remains the PR description table: source/depth/node/row/result/deadline refusal returns no payload; panic returns panic; native allocation/stack aborts are not contained. Tests: unchanged authored contracts, every exclusive counter/result boundary, seeded fuzz, depth and unwind recovery. If optimized larger buckets still fail, stop and report profile and numbers.

kayg changed title from ~~WIP: calendar v2 slice 2 blocked on A/C parser budgets~~ to WIP: calendar v2 slice 2 optimized in-process parser gates

2026-10-31 20:22:31 +00:00

kayg commented

2026-10-31 20:29:46 +00:00

Author

Owner

Independent review, round 1

Reviewed 75a1577f5b98909836146e3f3309a23932a6aca2. Verdict: fix first. P0: none.

Findings

P1: admitted short-line input can abort before returning unresolved. In crates/markdown-view/src/lines.rs:73–107, admission is followed by multiple whole-document per-line allocations without a working-memory reservation or intervening deadline checks. The line ceiling aliases the 8 MiB source ceiling, so it does not bound these allocations to the documented 256 MiB envelope. On dev-01, the existing in_process_containment release example with stdin b"\n" * 4194304, RLIMIT_AS=256 MiB, RLIMIT_STACK=8 MiB and core dumps disabled exits -6, reporting memory allocation of 67108880 bytes failed; no unresolved payload is returned. This input is below both source and line ceilings and has no extracted rows. A 1,048,576-newline probe returns timeout after 59.707583 ms. The report's maximum-size newline probe instead hits the line admission check, missing this intermediate allocation case. Add intermediate-size coverage and enforce a working-memory budget before allocating, or compact/stream the per-line state. This is a concrete instance of the acknowledged native-allocation limitation in selected option A, rather than a request to implement later application recovery.

P2: escaped tags and formatting boundaries violate the extraction contract and parser parity. At crates/markdown-view/src/lines.rs:149–177, the regex treats each Text-event source slice as a fresh metadata boundary. For - 09:00 Meet \#work, Rust returns tag #work at [14,19] and removes it from the title; the TS adapter returns no tags and preserves Meet \#work. Likewise - 09:00 Meet word**#tag** produces a Rust tag but no TS tag. The contract requires escaped tokens to remain literal and start/whitespace boundaries to survive formatting. Preserve escape/boundary context across events and add independently authored vectors to the shared rows corpus, exercised in both jobs. Both counterexamples were run against the existing native probe and the checked-in TS adapter with the pinned Lezer versions.

P2: the two CI jobs are not required merge checks. The workflow runs the shared classification, exact-span/source-slice and relationship expectations in both adapters, with no conditional skip in the row suites. However, live Forgejo API reads show dev.protected=false, enable_status_check=false, empty status_check_contexts, and no branch protections. Thus either job can be absent or red without enforcing the plan's mandatory gate. Configure required native/editor checks before treating the contract as non-skippable. No repository settings were changed during review.

Requested checks and limits

Composer: production options and title handling are unchanged. The native job runs the real tasks.create byte corpus in addition to the parser oracle; existing web CI discovers the added recogniser byte test. No consumer switches were found.
Pre-work limits: source/line admission precedes extraction and the vendored block-container depth check precedes opening the next container. Node/inline/emphasis checks occur only after OffsetIter::next(); Parser::new_ext_bounded constructs the full first-pass tree before the first event counter check. Result bytes are checked during serialization after rows and their strings have already been built. These are exact output/event ceilings, not complete pre-allocation work limits.
Deadline view: caps imply source <8 MiB, block-container depth <256, counted nodes/inlines/emphasis <250k, rows plus relationships <100k, and encoded result <16 MiB. They imply no demonstrated numeric worst-case wall time. Parser construction, individual iterator steps, allocation and cleanup can run outside clock checks; 64 events also need not represent a fixed amount of work. A cooperative deadline is adequate as a soft refusal mechanism for this unconnected experiment, but inadequate to establish a bounded 50 ms in-process service guarantee. The measured normal corpus cannot establish that bound.
Panics/state: the selected extract_in_process entry wraps extraction in catch_unwind, returns an error without a partial replacement, and exposes no FFI boundary. No publication state is mutated here. Retaining durable last-good state is a caller obligation, not demonstrated application behavior. Public raw extraction entry points and the historical worker do not themselves catch panics; any future integration must use the contained entry. Allocation/stack aborts remain outside unwind recovery.
Measurement: the direct paired baseline avoids charging IPC only to baseline; fixture byte/shard identity, two lanes, destruction inside the timer, and nearest-rank indexing are sound for the stated parser-only comparison. One ordered baseline/head pair supports the reported observation, not repeated application acceptance or a noise estimate. The benchmark prints budget booleans rather than failing on false, so it remains manually interpreted evidence.
Scope: no later-slice consumer, ledger, identity, source or DAV integration was found. Worker/pool/hybrid code remains unconnected historical experiment infrastructure. Application last-good lifecycle and acceptance gates remain deferred.

Verification

Control host: read-only code/API inspection; no builds, tests, installs or pushes. VM: existing SHA-verified native probe 265633da…5014, sacrificial in-process probes under the resource limits above, and nix develop -c bash -c '~/bin/slot bun -e …' for the three metadata examples. TS extraction SHA-256 matches the reviewed tree (0053023b…02de); parser versions are @codemirror/lang-markdown 6.5.2 and @lezer/markdown 1.8.0. Existing full gate results were inspected, not rerun. No code changed.

## Independent review, round 1 Reviewed `75a1577f5b98909836146e3f3309a23932a6aca2`. **Verdict: fix first.** P0: none. ### Findings **P1: admitted short-line input can abort before returning unresolved.** In `crates/markdown-view/src/lines.rs:73–107`, admission is followed by multiple whole-document per-line allocations without a working-memory reservation or intervening deadline checks. The line ceiling aliases the 8 MiB source ceiling, so it does not bound these allocations to the documented 256 MiB envelope. On dev-01, the existing `in_process_containment` release example with stdin `b"\n" * 4194304`, RLIMIT_AS=256 MiB, RLIMIT_STACK=8 MiB and core dumps disabled exits **-6**, reporting `memory allocation of 67108880 bytes failed`; no unresolved payload is returned. This input is below both source and line ceilings and has no extracted rows. A 1,048,576-newline probe returns timeout after **59.707583 ms**. The report's maximum-size newline probe instead hits the line admission check, missing this intermediate allocation case. Add intermediate-size coverage and enforce a working-memory budget before allocating, or compact/stream the per-line state. This is a concrete instance of the acknowledged native-allocation limitation in selected option A, rather than a request to implement later application recovery. **P2: escaped tags and formatting boundaries violate the extraction contract and parser parity.** At `crates/markdown-view/src/lines.rs:149–177`, the regex treats each Text-event source slice as a fresh metadata boundary. For `- 09:00 Meet \#work`, Rust returns tag `#work` at [14,19] and removes it from the title; the TS adapter returns no tags and preserves `Meet \#work`. Likewise `- 09:00 Meet word**#tag**` produces a Rust tag but no TS tag. The contract requires escaped tokens to remain literal and start/whitespace boundaries to survive formatting. Preserve escape/boundary context across events and add independently authored vectors to the shared rows corpus, exercised in both jobs. Both counterexamples were run against the existing native probe and the checked-in TS adapter with the pinned Lezer versions. **P2: the two CI jobs are not required merge checks.** The workflow runs the shared classification, exact-span/source-slice and relationship expectations in both adapters, with no conditional skip in the row suites. However, live Forgejo API reads show `dev.protected=false`, `enable_status_check=false`, empty `status_check_contexts`, and no branch protections. Thus either job can be absent or red without enforcing the plan's mandatory gate. Configure required native/editor checks before treating the contract as non-skippable. No repository settings were changed during review. ### Requested checks and limits - **Composer:** production options and title handling are unchanged. The native job runs the real tasks.create byte corpus in addition to the parser oracle; existing web CI discovers the added recogniser byte test. No consumer switches were found. - **Pre-work limits:** source/line admission precedes extraction and the vendored block-container depth check precedes opening the next container. Node/inline/emphasis checks occur only after `OffsetIter::next()`; `Parser::new_ext_bounded` constructs the full first-pass tree before the first event counter check. Result bytes are checked during serialization after rows and their strings have already been built. These are exact output/event ceilings, not complete pre-allocation work limits. - **Deadline view:** caps imply source <8 MiB, block-container depth <256, counted nodes/inlines/emphasis <250k, rows plus relationships <100k, and encoded result <16 MiB. They imply no demonstrated numeric worst-case wall time. Parser construction, individual iterator steps, allocation and cleanup can run outside clock checks; 64 events also need not represent a fixed amount of work. A cooperative deadline is adequate as a soft refusal mechanism for this unconnected experiment, but inadequate to establish a bounded 50 ms in-process service guarantee. The measured normal corpus cannot establish that bound. - **Panics/state:** the selected `extract_in_process` entry wraps extraction in `catch_unwind`, returns an error without a partial replacement, and exposes no FFI boundary. No publication state is mutated here. Retaining durable last-good state is a caller obligation, not demonstrated application behavior. Public raw extraction entry points and the historical worker do not themselves catch panics; any future integration must use the contained entry. Allocation/stack aborts remain outside unwind recovery. - **Measurement:** the direct paired baseline avoids charging IPC only to baseline; fixture byte/shard identity, two lanes, destruction inside the timer, and nearest-rank indexing are sound for the stated parser-only comparison. One ordered baseline/head pair supports the reported observation, not repeated application acceptance or a noise estimate. The benchmark prints budget booleans rather than failing on false, so it remains manually interpreted evidence. - **Scope:** no later-slice consumer, ledger, identity, source or DAV integration was found. Worker/pool/hybrid code remains unconnected historical experiment infrastructure. Application last-good lifecycle and acceptance gates remain deferred. ### Verification Control host: read-only code/API inspection; no builds, tests, installs or pushes. VM: existing SHA-verified native probe `265633da…5014`, sacrificial in-process probes under the resource limits above, and `nix develop -c bash -c '~/bin/slot bun -e …'` for the three metadata examples. TS extraction SHA-256 matches the reviewed tree (`0053023b…02de`); parser versions are @codemirror/lang-markdown 6.5.2 and @lezer/markdown 1.8.0. Existing full gate results were inspected, not rerun. No code changed.

kayg commented

2026-10-31 20:40:50 +00:00

Author

Owner

Round 1 progress: reproduced the 4 MiB newline allocation abort under 256 MiB AS / 8 MiB stack. Added memory-limit regression, raw-byte admission caps and a shared extraction allocation counter. Both metadata counterexamples now have independently authored shared spans and pass native/editor contracts (201 editor tests). Native tests, clippy -D warnings and editor typecheck pass. The first full-shape 4 MiB memory sweep stayed below 24 MiB peak RSS, including admission-edge tags/links. The first paired perf run fails the unchanged speed budget; profiling/optimization is in progress and no speed pass is claimed. Forgejo dev protection now requires Markdown extraction contract / markdown-native-contract (pull_request) and Markdown extraction contract / markdown-editor-contract (pull_request). Hosted contexts are currently absent, which blocks merge. Never merge.

kayg commented

2026-10-31 20:49:53 +00:00

Author

Owner

Round 1 implementation status (max 3 rounds)

Head: 343acaa60bab1995b7ed450865c506a9bbdfdc0e. Never merge.

item	status	evidence
P1 memory reproduction	PASS	Before: 4 MiB newlines exit -6 under 256 MiB AS / 8 MiB stack. After: unresolved/blank-run-limit; 1 MiB also refused safely. CI repeats the 4 MiB case under ulimit.
Memory admission/allocation	PASS requested shape matrix	Raw-byte size/line/blank-run scan and conservative parser reservation precede parser/per-line allocation. Adapter capacities and source copies charge a shared 64 MiB counter before allocation.
4 MiB memory edges	PASS	Blank lines 19,624 KiB; depth-255 lists 23,768 KiB; long event line 18,336 KiB; tags density search 24,612 KiB; links search 23,788 KiB peak RSS. Maximum 24.04 MiB; separate final 32 MiB regression guard passes.
P2 metadata parity	PASS	Two independently authored shared vectors preserve escaped tags and word/formatting boundaries. All old expectations unchanged; native corpus and 201 editor tests pass.
P2 required contract checks	CONFIGURED	dev protection requires both exact native/editor pull_request contexts, including admins. Both workflow steps use set -eu; no continue-on-error.
Native gates	PASS	22 crate tests, original 70/70 Rust line vectors, fmt, clippy -D warnings; 12 Composer create tests.
Editor gates	PASS	Locked install, editor typecheck, 201 tests across 3 files.
Speed budget	PASS	4/32/256 KiB p95 growth +7.71% / -2.25% / -9.10%; added p95 +0.003747 / -0.007935 / -0.288992 ms. 100k wall 5.458378s direct vs 5.540171s head (+1.50%).
Hosted CI	ABSENT / MERGE BLOCKED	Required contexts have no hosted status on this head. VM passes are not presented as hosted statuses.

Decisions: source/line length <=4 MiB, physical lines <=16,384, consecutive whitespace-only lines <=1,024; 64 MiB exclusive working-byte accounting ceiling. Parser reservation is 64 KiB + 4*bytes + 256*lines + 256*ASCII punctuation; all adapter growth/copies are additional charges. Complexity remains linear in source, events and encoded output. The 50 ms deadline remains cooperative. Memory-only probes relax time to 30s so cleanup/full admitted allocation is measured with production memory caps unchanged. Native allocation/stack aborts are not unwind containment; unrelated process allocations/runtime overhead are outside the reservation model. No consumer or publication integration.

The first round-1 speed run failed (+28.22% 100k), and the table-only scan still failed the 4 KiB gate (+18.28%). Profiling identified punctuation range checks and lost event-loop inlining; final table counting plus an explicitly inline event step restored all budgets. Both earlier failures remain documented. Benchmark budget booleans now also produce a failing exit status; assertions were verified against the final recorded reports.

Same original direct baseline, seed-805 byte identity for 100 variants, original two shard hashes, two lanes, warm 10k then 100k, nearest-rank quantiles. Zero unresolved, 925,696,000 bytes, 200k rows, 50k relationships. Head release SHA-256 verified on dev-01, relay and perf-02: parser_options 10dcb1de94b0d747e6f37eb7331c6100519c88ffff45030d37a1de458ee34bc3; memory probe 939ae4fb19bcab766599ecc1247726332864727145d8b0c03481f42efe5b98d7.

Exact dev-01 gates, in Nix and each heavy command via ~/bin/slot:

cargo test -j4 -p calternal-markdown-view --locked
cargo test -j4 --locked --manifest-path docs/plans/0067-vectors/rust/Cargo.toml
cargo fmt -p calternal-markdown-view --check
cargo clippy -j4 -p calternal-markdown-view --all-targets --locked -- -D warnings
cargo test -j4 -p calternal-tasks --locked create_tests::
bun install --frozen-lockfile
bun run --cwd packages/editor check
# From packages/editor:
bunx vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts

Perf-02:

~/bin/perf-lock python3 ~/calendar805-r1/check-markdown-memory.py ~/calendar805-r1/in_process_containment
~/bin/perf-lock python3 ~/calendar805-r1/compare-optimized.py ~/calendar805-r1

Memory model, failure table, exact measurements, hashes and historical evidence. Raw final logs: dev-01 ~/calendar805-r1-{last-gates,final,verify}.log; perf-02 ~/calendar805-r1-{memory-final,memory-guard,perf3}.log. Every meaningful step committed/pushed. No control-host builds/tests/installs; no merge.

## Round 1 implementation status (max 3 rounds) Head: `343acaa60bab1995b7ed450865c506a9bbdfdc0e`. **Never merge.** | item | status | evidence | | --- | --- | --- | | P1 memory reproduction | PASS | Before: 4 MiB newlines exit -6 under 256 MiB AS / 8 MiB stack. After: unresolved/blank-run-limit; 1 MiB also refused safely. CI repeats the 4 MiB case under ulimit. | | Memory admission/allocation | PASS requested shape matrix | Raw-byte size/line/blank-run scan and conservative parser reservation precede parser/per-line allocation. Adapter capacities and source copies charge a shared 64 MiB counter before allocation. | | 4 MiB memory edges | PASS | Blank lines 19,624 KiB; depth-255 lists 23,768 KiB; long event line 18,336 KiB; tags density search 24,612 KiB; links search 23,788 KiB peak RSS. Maximum 24.04 MiB; separate final 32 MiB regression guard passes. | | P2 metadata parity | PASS | Two independently authored shared vectors preserve escaped tags and word/formatting boundaries. All old expectations unchanged; native corpus and 201 editor tests pass. | | P2 required contract checks | CONFIGURED | dev protection requires both exact native/editor pull_request contexts, including admins. Both workflow steps use set -eu; no continue-on-error. | | Native gates | PASS | 22 crate tests, original 70/70 Rust line vectors, fmt, clippy -D warnings; 12 Composer create tests. | | Editor gates | PASS | Locked install, editor typecheck, 201 tests across 3 files. | | Speed budget | PASS | 4/32/256 KiB p95 growth +7.71% / -2.25% / -9.10%; added p95 +0.003747 / -0.007935 / -0.288992 ms. 100k wall 5.458378s direct vs 5.540171s head (+1.50%). | | Hosted CI | ABSENT / MERGE BLOCKED | Required contexts have no hosted status on this head. VM passes are not presented as hosted statuses. | Decisions: source/line length <=4 MiB, physical lines <=16,384, consecutive whitespace-only lines <=1,024; 64 MiB exclusive working-byte accounting ceiling. Parser reservation is `64 KiB + 4*bytes + 256*lines + 256*ASCII punctuation`; all adapter growth/copies are additional charges. Complexity remains linear in source, events and encoded output. The 50 ms deadline remains cooperative. Memory-only probes relax time to 30s so cleanup/full admitted allocation is measured with production memory caps unchanged. Native allocation/stack aborts are not unwind containment; unrelated process allocations/runtime overhead are outside the reservation model. No consumer or publication integration. The first round-1 speed run failed (+28.22% 100k), and the table-only scan still failed the 4 KiB gate (+18.28%). Profiling identified punctuation range checks and lost event-loop inlining; final table counting plus an explicitly inline event step restored all budgets. Both earlier failures remain documented. Benchmark budget booleans now also produce a failing exit status; assertions were verified against the final recorded reports. Same original direct baseline, seed-805 byte identity for 100 variants, original two shard hashes, two lanes, warm 10k then 100k, nearest-rank quantiles. Zero unresolved, 925,696,000 bytes, 200k rows, 50k relationships. Head release SHA-256 verified on dev-01, relay and perf-02: parser_options `10dcb1de94b0d747e6f37eb7331c6100519c88ffff45030d37a1de458ee34bc3`; memory probe `939ae4fb19bcab766599ecc1247726332864727145d8b0c03481f42efe5b98d7`. Exact dev-01 gates, in Nix and each heavy command via ~/bin/slot: ```sh cargo test -j4 -p calternal-markdown-view --locked cargo test -j4 --locked --manifest-path docs/plans/0067-vectors/rust/Cargo.toml cargo fmt -p calternal-markdown-view --check cargo clippy -j4 -p calternal-markdown-view --all-targets --locked -- -D warnings cargo test -j4 -p calternal-tasks --locked create_tests:: bun install --frozen-lockfile bun run --cwd packages/editor check # From packages/editor: bunx vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts ``` Perf-02: ```sh ~/bin/perf-lock python3 ~/calendar805-r1/check-markdown-memory.py ~/calendar805-r1/in_process_containment ~/bin/perf-lock python3 ~/calendar805-r1/compare-optimized.py ~/calendar805-r1 ``` [Memory model, failure table, exact measurements, hashes and historical evidence](https://git.kayg.org/kayg/calternal/src/branch/feat/calendar-v2-s2-parser-contract/docs/plans/0067-slice2/README.md). Raw final logs: dev-01 ~/calendar805-r1-{last-gates,final,verify}.log; perf-02 ~/calendar805-r1-{memory-final,memory-guard,perf3}.log. Every meaningful step committed/pushed. No control-host builds/tests/installs; no merge.

kayg commented

2026-10-31 20:51:00 +00:00

Author

Owner

Orchestrator note: the dev branch-protection rule created at 2026-10-31 20:36 UTC (required status checks for the two Markdown contract jobs) was removed. It was outside this PR's scope, and Forgejo doesn't report Actions statuses here, so it blocked all merges into dev. "Required" for this PR means the workflow jobs fail the workflow (no continue-on-error), which is already in the branch. Repo settings stay with the orchestrator and kayg.

Orchestrator note: the `dev` branch-protection rule created at 2026-10-31 20:36 UTC (required status checks for the two Markdown contract jobs) was removed. It was outside this PR's scope, and Forgejo doesn't report Actions statuses here, so it blocked all merges into dev. "Required" for this PR means the workflow jobs fail the workflow (no continue-on-error), which is already in the branch. Repo settings stay with the orchestrator and kayg.

kayg commented

2026-10-31 20:55:44 +00:00

Author

Owner

Independent review, round 2

Reviewed 343acaa60bab1995b7ed450865c506a9bbdfdc0e. Verdict: fix first. New findings: P0 0, P1 1, P2 1, P3 0. No code pushed or repository settings changed.

Round-1 status

Round-1 item	Status	Does its test demonstrate the fix?
P1: intermediate short-line input aborts before unresolved	not resolved overall; original LF reproduction fixed	Yes for the reported input: 4 MiB LF now returns `unresolved/blank-run-limit`, and CI checks that outcome under 256 MiB AS / 8 MiB stack. No for the general memory guarantee: bare CR passes admission and still aborts (P1 below).
P2: escaped tags and formatting boundaries	resolved for both reported cases; parity regressed elsewhere	The two new independently authored shared vectors exercise the exact escaped-tag and word-before-formatting examples in both adapters. Both pass. They omit the adjacent formatting/link cases below, which now disagree.
P2: contract CI jobs must fail the workflow	resolved under the clarified scope	Both jobs run on pull requests, have no `continue-on-error` or conditional contract skip, and their multi-command contract steps use `set -eu`. Test failures propagate to job/workflow failure. This is workflow inspection, not a hosted-run claim; branch protection is outside this review.

New findings

P1: bare-CR lines bypass admission and parser reservation, retaining the allocation-abort path.

In crates/markdown-view/src/bounded.rs:97–118, admission uses source.split('\n'); CR is neither a line separator for the line/blank-run counters nor ASCII punctuation for the syntax reservation. Pulldown explicitly accepts bare CR as an end-of-line (vendor/pulldown-cmark/src/scanners.rs:493–499). Consequently b"x\r" * 2_097_152 is admitted as one 4 MiB line with zero punctuation and a parser reservation of only 16,843,008 bytes, while the parser builds state for millions of lines before event checks.

Fresh reproduction on dev-01, using the report's SHA-256-matched release in_process_containment binary (939ae4fb…5b98d7), production deadline, 256 MiB AS / 8 MiB stack / no core dumps:

import resource, subprocess
def caps():
    resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_CORE, (0, 0))
p = subprocess.run(
    ["target/release/examples/in_process_containment"],
    input=b"x\r" * 2_097_152, capture_output=True,
    preexec_fn=caps, timeout=15)
print(p.returncode, p.stdout, p.stderr)

Result: exit -6, empty stdout, memory allocation of 402653184 bytes failed. No unresolved result is returned. This is the same resource envelope as round 1. The adapter's explicit arrays, vector growth and copied spans are charged, but the parser's uninstrumented allocations rely on the reservation, so this admission mismatch bypasses that model.

Align admission with every line ending the parser accepts, or refuse unsupported line endings before parser construction; preserve original source offsets. Add bare-CR and mixed LF/CRLF/CR intermediate-size and cap-edge cases to the resource-limited regression. The existing long-line, LF blank-run, depth-255, tag-density and link-density measurements are useful, but do not cover this shape, so their 24.04 MiB maximum/32 MiB guard cannot establish the general envelope.

P2: the new native boundary check creates additional native/editor metadata disagreements.

The added check at crates/markdown-view/src/lines.rs:177–192 retains preceding source context, while the editor matcher at packages/editor/src/markdown/extraction.ts:108–112 still treats each visible fragment as a fresh start. Fresh probes against the installed editor parser and native adapter give:

Source	Native tags	Editor tags
`- 09:00 Meet word#tag`	`[]`	`["#tag"]`
`- 09:00 Meet word#tag`	`[]`	`["#tag"]`
`- 09:00 Meet [word](p)#tag`	`[]`	`["#tag"]`
`- 09:00 Meet word[#tag](p)`	`[]`	`["#tag"]`

The native fix now rejects these tags; the unchanged editor path still extracts them and removes them from the title. The start/whitespace rule needs consistent source context on both sides. Add independently authored shared vectors for these cases and retain the two existing fix vectors unchanged.

Cooperative deadline

A cooperative deadline is acceptable as a soft refusal mechanism for this unconnected experimental slice, once the admission/resource defect is fixed. It does not demonstrate a 50 ms worst-case bound. Parser construction, a single iterator step, scans and cleanup remain outside clock checkpoints; capped bytes/events alone do not supply a measured numeric wall-time upper bound.

A fresh b"\r" * 4_194_304 production-deadline probe returned timeout after 105.057715 ms on dev-01. The bare-CR allocation abort above also occurs before a timeout result. The normal-corpus speed pass and the five measured memory shapes therefore cannot establish worst-case time. Keep the soft-deadline limitation explicit and include all accepted line-ending shapes in the next edge measurements; this head is fix first because of the concrete allocation and parity findings, not because this slice must introduce a worker.

Verification

All execution was on dev-01 in Nix, through ~/bin/slot; no control-host builds/tests/installs. The existing VM worktree has an older commit label with modified files: reviewed bounded.rs, lines.rs and vendored firstpass.rs blob hashes match 343acaa6, and the release probe hash matches the recorded artifact.

Fresh passes:

cargo test -j4 -p calternal-markdown-view --locked: 22 passed.
From packages/editor: bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts: 201 passed, 3 files.
Direct native/editor probes confirm both original metadata counterexamples are fixed and the four disagreements above remain.
Resource-limited LF/CR/NULL probes: original 4 MiB LF safely refused; 4 MiB NULL accepted; bare-CR results described above.

An initial Vitest invocation from the repository root selected an external Vitest and failed before running tests because happy-dom was unavailable there. The corrected command above used the installed editor Vitest 5.0.4 and passed. Existing Composer/typecheck/clippy and perf-02 memory/speed evidence was inspected, not rerun; no fresh whole-application or worst-case-time pass is claimed.

## Independent review, round 2 Reviewed `343acaa60bab1995b7ed450865c506a9bbdfdc0e`. **Verdict: fix first.** New findings: P0 0, P1 1, P2 1, P3 0. No code pushed or repository settings changed. ### Round-1 status | Round-1 item | Status | Does its test demonstrate the fix? | | --- | --- | --- | | P1: intermediate short-line input aborts before unresolved | **not resolved overall**; original LF reproduction fixed | Yes for the reported input: 4 MiB LF now returns `unresolved/blank-run-limit`, and CI checks that outcome under 256 MiB AS / 8 MiB stack. No for the general memory guarantee: bare CR passes admission and still aborts (P1 below). | | P2: escaped tags and formatting boundaries | **resolved for both reported cases; parity regressed elsewhere** | The two new independently authored shared vectors exercise the exact escaped-tag and word-before-formatting examples in both adapters. Both pass. They omit the adjacent formatting/link cases below, which now disagree. | | P2: contract CI jobs must fail the workflow | **resolved under the clarified scope** | Both jobs run on pull requests, have no `continue-on-error` or conditional contract skip, and their multi-command contract steps use `set -eu`. Test failures propagate to job/workflow failure. This is workflow inspection, not a hosted-run claim; branch protection is outside this review. | ### New findings **P1: bare-CR lines bypass admission and parser reservation, retaining the allocation-abort path.** In `crates/markdown-view/src/bounded.rs:97–118`, admission uses `source.split('\n')`; CR is neither a line separator for the line/blank-run counters nor ASCII punctuation for the syntax reservation. Pulldown explicitly accepts bare CR as an end-of-line (`vendor/pulldown-cmark/src/scanners.rs:493–499`). Consequently `b"x\r" * 2_097_152` is admitted as one 4 MiB line with zero punctuation and a parser reservation of only **16,843,008 bytes**, while the parser builds state for millions of lines before event checks. Fresh reproduction on dev-01, using the report's SHA-256-matched release `in_process_containment` binary (`939ae4fb…5b98d7`), production deadline, 256 MiB AS / 8 MiB stack / no core dumps: ```python import resource, subprocess def caps(): resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_CORE, (0, 0)) p = subprocess.run( ["target/release/examples/in_process_containment"], input=b"x\r" * 2_097_152, capture_output=True, preexec_fn=caps, timeout=15) print(p.returncode, p.stdout, p.stderr) ``` Result: **exit -6**, empty stdout, `memory allocation of 402653184 bytes failed`. No unresolved result is returned. This is the same resource envelope as round 1. The adapter's explicit arrays, vector growth and copied spans are charged, but the parser's uninstrumented allocations rely on the reservation, so this admission mismatch bypasses that model. Align admission with every line ending the parser accepts, or refuse unsupported line endings before parser construction; preserve original source offsets. Add bare-CR and mixed LF/CRLF/CR intermediate-size and cap-edge cases to the resource-limited regression. The existing long-line, LF blank-run, depth-255, tag-density and link-density measurements are useful, but do not cover this shape, so their 24.04 MiB maximum/32 MiB guard cannot establish the general envelope. **P2: the new native boundary check creates additional native/editor metadata disagreements.** The added check at `crates/markdown-view/src/lines.rs:177–192` retains preceding source context, while the editor matcher at `packages/editor/src/markdown/extraction.ts:108–112` still treats each visible fragment as a fresh start. Fresh probes against the installed editor parser and native adapter give: | Source | Native tags | Editor tags | | --- | --- | --- | | `- 09:00 Meet *word*#tag` | `[]` | `["#tag"]` | | `- 09:00 Meet **word**#tag` | `[]` | `["#tag"]` | | `- 09:00 Meet [word](p)#tag` | `[]` | `["#tag"]` | | `- 09:00 Meet word[#tag](p)` | `[]` | `["#tag"]` | The native fix now rejects these tags; the unchanged editor path still extracts them and removes them from the title. The start/whitespace rule needs consistent source context on both sides. Add independently authored shared vectors for these cases and retain the two existing fix vectors unchanged. ### Cooperative deadline A cooperative deadline is acceptable as a **soft refusal mechanism for this unconnected experimental slice**, once the admission/resource defect is fixed. It does not demonstrate a 50 ms worst-case bound. Parser construction, a single iterator step, scans and cleanup remain outside clock checkpoints; capped bytes/events alone do not supply a measured numeric wall-time upper bound. A fresh `b"\r" * 4_194_304` production-deadline probe returned timeout after **105.057715 ms** on dev-01. The bare-CR allocation abort above also occurs before a timeout result. The normal-corpus speed pass and the five measured memory shapes therefore cannot establish worst-case time. Keep the soft-deadline limitation explicit and include all accepted line-ending shapes in the next edge measurements; this head is **fix first** because of the concrete allocation and parity findings, not because this slice must introduce a worker. ### Verification All execution was on dev-01 in Nix, through `~/bin/slot`; no control-host builds/tests/installs. The existing VM worktree has an older commit label with modified files: reviewed `bounded.rs`, `lines.rs` and vendored `firstpass.rs` blob hashes match `343acaa6`, and the release probe hash matches the recorded artifact. Fresh passes: - `cargo test -j4 -p calternal-markdown-view --locked`: **22 passed**. - From `packages/editor`: `bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts`: **201 passed**, 3 files. - Direct native/editor probes confirm both original metadata counterexamples are fixed and the four disagreements above remain. - Resource-limited LF/CR/NULL probes: original 4 MiB LF safely refused; 4 MiB NULL accepted; bare-CR results described above. An initial Vitest invocation from the repository root selected an external Vitest and failed before running tests because `happy-dom` was unavailable there. The corrected command above used the installed editor Vitest 5.0.4 and passed. Existing Composer/typecheck/clippy and perf-02 memory/speed evidence was inspected, not rerun; no fresh whole-application or worst-case-time pass is claimed.

kayg commented

2026-10-31 21:20:17 +00:00

Author

Owner

Round 2 implementation status

Head: 948117ccc9dbcfa02e044dc5efc6484a32f414e1. Never merge. No repository settings changed. Ready for the independent final review.

Item	Status	Evidence
P1 bare-CR reproduction	PASS	Before: `x\r` × 2,097,152 exits -6 with allocation failure under 256 MiB AS / 8 MiB stack. After: unresolved/line-limit. Intermediate 1 MiB and 4 MiB CR, LF, CRLF and mixed cases safely refuse.
One normalized line view	PASS	One allocation-free iterator defines LF, CRLF and lone CR as breaks, including a final empty line. Every line/blank-run/length/syntax admission count, canonical LF materialization and walk line array derives from it. Parser ranges map back to original byte offsets before extraction. Text and offset-map growth charge the shared budget first.
Memory property / cap edges	PASS	72 seeded independent Linux child cases in crate tests, real 256 MiB AS / 8 MiB stack, 32 MiB VmHWM guard; limit setup must succeed and both accepted/refused cases must occur. Separately: 136 release probes, including 72 generated mixed-ending/blank-run/nesting cases, cap edges and density searches; peak 30,052 KiB, below unchanged 32 MiB RSS guard.
P2 four metadata disagreements	PASS	Four independently authored shared vectors pass both adapters with the editor's existing decisions: #tag is extracted after word, word, word, and inside word#tag. Both round-1 fix vectors are unchanged.
Original offsets / line semantics	PASS	Shared Unicode/mixed-ending/frontmatter/task/child-link vectors check exact spans and source slices in both adapters. Lone CR always breaks a line, so the old title-trim-0xd vector intentionally changes to three lines with an empty first-event title.
Native gates	PASS	25 crate tests; original 70 Rust line vectors; 12 Composer create tests; fmt; clippy -D warnings.
Editor gates	PASS	Locked install, typecheck, 207 tests across 3 contract files.
CI memory regression	PASS on VM	Executed the exact workflow LF/bare-CR/mixed shell probes under ulimit, asserting unresolved reasons. Native crate property also runs in the native CI job. No conditional contract skip or continue-on-error.
Speed budget	PASS	4/32/256 KiB p95 growth +7.31% / -2.14% / -9.69%; added p95 +0.003538 / -0.007543 / -0.315301 ms. Paired 100k wall 5.457216405s direct / 5.526577250s head (+1.27%).
Hosted checks / final review	NOT CLAIMED	Forgejo reports no commit status contexts; VM gates are separate evidence. Independent final review remains pending.

577d893232a852172c514f0cab37b5d89c1a7509d1616575eb68730fb8869c65 parser_options
d95600e5c25ba09c2c50faa38631b170d6e47e8ecfb80d44e358d32f90f36229 in_process_containment

Exact dev-01 gates, in Nix, each heavy command through ~/bin/slot:

cargo test -j4 -p calternal-markdown-view --locked
cargo test -j4 --locked --manifest-path docs/plans/0067-vectors/rust/Cargo.toml
cargo fmt -p calternal-markdown-view --check
cargo clippy -j4 -p calternal-markdown-view --all-targets --locked -- -D warnings
cargo test -j4 -p calternal-tasks --locked create_tests::
cargo build -j4 --release -p calternal-markdown-view --examples --locked
bun install --frozen-lockfile
bun run --cwd packages/editor check
# From packages/editor:
bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts

Perf-02, under perf-lock:

~/bin/perf-lock python3 ~/calendar805-r2/check-markdown-memory.py ~/calendar805-r2/in_process_containment
~/bin/perf-lock python3 ~/calendar805-r2/compare-optimized.py ~/calendar805-r2

Every meaningful step committed and pushed. No control-host builds/tests/installs; no merge.

kayg referenced this pull request

2026-10-31 21:24:00 +00:00

feat(tasks): port B1 status plans to the Home executor #812

kayg commented

2026-10-31 21:27:02 +00:00

Author

Owner

Independent review, round 3 (final)

Reviewed 948117ccc9dbcfa02e044dc5efc6484a32f414e1. Verdict: small fix (P2 only). New findings: P0 0, P1 0, P2 2, P3 0. No code pushed, merge performed or repository settings changed.

Item	Status	Independent evidence
Round-2 P1: bare CR bypasses admission and aborts	resolved	Fresh `b"x\r" * 2_097_152` probe under real 256 MiB AS / 8 MiB stack / no core dumps: exit 0, `unresolved/line-limit`, 0.738123 ms with the production deadline; also refused with the 30s measurement deadline.
Admission and extraction line accounting	resolved	`Budget::new`, canonical LF construction and extraction arrays all derive from `source::lines(source)`. LF, CRLF and lone CR share one separator implementation and final-empty-line rule; the parser receives canonical LF, never the raw CR stream. Pre-scan and array population therefore cannot count different lines for the same source. All 120 additional mixed-ending comparisons passed, including Unicode spans, frontmatter, tasks and child links.
Intermediate-size and cap-edge memory checks	pass for tested shapes	16 fresh limited release runs: old LF reproduction, bare CR, mixed endings, blank CR, 16,384-line CR/CRLF edges, 4 MiB long text and long text ending in CR; production and 30s modes, every exit 0. Fresh crate run also passed the 72-child mixed-ending/blank-run/nesting property under its real limits and 32 MiB VmHWM guard. These are regression evidence, not a universal allocator proof.
Round-2 P2: four named metadata disagreements	resolved for all four	Fresh native/editor probes agree after `word`, `word`, `[word](p)` and inside `word[#tag](p)`. Both round-1 examples still agree: escaped tags and `word#tag` remain literal. Additional P2 cases below remain.
Native/editor contracts	pass	Fresh: 25 Rust tests; 207 editor tests across 3 files. All 13 runtime/contract/gate files in the reported manifest independently match this head. Release containment binary SHA-256 is `d95600e5c25ba09c2c50faa38631b170d6e47e8ecfb80d44e358d32f90f36229`.
Composer, fmt, clippy, typecheck and speed	prior evidence inspected	Existing final logs record 12 Composer tests, fmt, clippy and editor typecheck passing. Checked-in paired speed evidence reports p95 growth +7.31% / -2.14% / -9.69%, paired 100k wall +1.27%. These gates were not independently rerun in this review. Workflow failures propagate; no hosted-check pass is claimed.

New findings

P2: native metadata still depends on inline event boundaries that differ from the editor's visible fragments.

crates/markdown-view/src/lines.rs:172–190,223–236 resets a fragment at a parsed link label and wrapper end, but not at a subscript/superscript start. It also matches each Text event separately, so a literal character split across Text events can truncate a tag. Fresh probes against the rebuilt native example and installed editor parser give:

Source	Native tags	Editor tags
`- 09:00 Meet word~#tag~`	`[]`	`#tag` at [18,22]
`- 09:00 Meet word^#tag^`	`[]`	`#tag` at [18,22]
`- 09:00 Meet #tag_#tag`	`#tag` at [15,19]	`#tag_` at [15,20]

The first two native titles retain the whole Meet word~#tag~ / Meet word^#tag^; the editor returns Meet word. The third case assigns the underscore to different fields. Match the editor's actual fragment boundaries, including opening wrappers and adjacent literal Text events; add independent shared vectors while retaining all six earlier repair vectors. A 675-case ASCII composition sweep found further disagreements in this area; the table gives minimal independently rechecked examples, not a count of distinct defects.

P2: the editor can remove escaped literal content when removing a tag-only wrapper.

At packages/editor/src/markdown/extraction.ts:73–87,145–158, collect skips non-inline child nodes, then onlyTags checks only the remaining visible gaps. An Escape can therefore be ignored when deciding that the entire wrapper contains metadata only.

For source - 09:00 Meet *#tag \** (one literal backslash before the first closing star), both adapters extract #tag at [14,18]. The editor returns title Meet, removing the escaped literal star too. Native retains title source slices Meet * and \**. The literal star is not metadata and must survive. Include literal/escaped children when deciding whether a wrapper contains only tags, and add a shared exact-title/source-span regression.

Cooperative deadline and worst-case time

The effective admission caps are source ≤4 MiB, logical lines ≤16,384, consecutive blank lines ≤1,024, and block-container depth <256. Node/inline/emphasis counters are each <250,000, rows plus relationships <100,000, encoded result <16 MiB, and the working reservation <64 MiB. The reservation models parser memory; it is not a process-wide allocator limit.

These caps bound input and counted output/work dimensions. They do not establish a numeric worst-case elapsed-time bound, including 50 ms. Admission can scan a long line between checks; normalization, full first-pass parser construction, an individual iterator step, wrapper/tag work and destruction can run outside clock checks. A group of 64 events has no fixed time cost. Offset remapping adds logarithmic endpoint lookup when CRLF is present. The observable return time can exceed the deadline by unchecked work and cleanup, plus scheduling; that excess has no demonstrated numeric maximum here.

The 50 ms cooperative deadline is acceptable as a soft refusal mechanism for this unconnected experimental slice. The caps, memory samples and normal-corpus speed pass do not justify a hard 50 ms in-process service guarantee. The current documentation states this limitation. No new P0/P1 was found, so the round-3 stop-and-simplify condition is not triggered.

Verification

All execution was on dev-01, in Nix through ~/bin/slot, using the existing calendar-805-r2 worktree and its configured shared target directory. Its commit label is older, but all 13 reviewed runtime/contract/gate files match this head byte-for-byte. Control host: read-only repository/API inspection and review preparation; no builds, tests or installs.

Fresh commands (with CARGO_TARGET_DIR=/home/claude/Developer/Worktrees/calendar-805-r1/target):

cargo test -j4 -p calternal-markdown-view --locked
cargo build -j4 --release -p calternal-markdown-view --example native_gate_probe --locked
# packages/editor:
bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts

The limited Python probes used resource.setrlimit before exec: RLIMIT_AS=(268435456,268435456), RLIMIT_STACK=(8388608,8388608), RLIMIT_CORE=(0,0), then subprocess.run([in_process_containment], input=b"x\r"*2097152, capture_output=True, preexec_fn=caps, timeout=35); the second lane adds --memory. Native/editor comparisons used the rebuilt native_gate_probe and extract(source, markdownLanguage.parser.parse(normalizeSource(source).slice(bodyOffset(normalizeSource(source))))), converting editor UTF-16 ranges to original UTF-8 byte ranges.

## Independent review, round 3 (final) Reviewed `948117ccc9dbcfa02e044dc5efc6484a32f414e1`. **Verdict: small fix (P2 only).** New findings: P0 0, P1 0, P2 2, P3 0. No code pushed, merge performed or repository settings changed. | Item | Status | Independent evidence | | --- | --- | --- | | Round-2 P1: bare CR bypasses admission and aborts | **resolved** | Fresh `b"x\r" * 2_097_152` probe under real 256 MiB AS / 8 MiB stack / no core dumps: exit 0, `unresolved/line-limit`, 0.738123 ms with the production deadline; also refused with the 30s measurement deadline. | | Admission and extraction line accounting | **resolved** | `Budget::new`, canonical LF construction and extraction arrays all derive from `source::lines(source)`. LF, CRLF and lone CR share one separator implementation and final-empty-line rule; the parser receives canonical LF, never the raw CR stream. Pre-scan and array population therefore cannot count different lines for the same source. All 120 additional mixed-ending comparisons passed, including Unicode spans, frontmatter, tasks and child links. | | Intermediate-size and cap-edge memory checks | **pass for tested shapes** | 16 fresh limited release runs: old LF reproduction, bare CR, mixed endings, blank CR, 16,384-line CR/CRLF edges, 4 MiB long text and long text ending in CR; production and 30s modes, every exit 0. Fresh crate run also passed the 72-child mixed-ending/blank-run/nesting property under its real limits and 32 MiB VmHWM guard. These are regression evidence, not a universal allocator proof. | | Round-2 P2: four named metadata disagreements | **resolved for all four** | Fresh native/editor probes agree after `*word*`, `**word**`, `[word](p)` and inside `word[#tag](p)`. Both round-1 examples still agree: escaped tags and `word**#tag**` remain literal. Additional P2 cases below remain. | | Native/editor contracts | **pass** | Fresh: 25 Rust tests; 207 editor tests across 3 files. All 13 runtime/contract/gate files in the reported manifest independently match this head. Release containment binary SHA-256 is `d95600e5c25ba09c2c50faa38631b170d6e47e8ecfb80d44e358d32f90f36229`. | | Composer, fmt, clippy, typecheck and speed | **prior evidence inspected** | Existing final logs record 12 Composer tests, fmt, clippy and editor typecheck passing. Checked-in paired speed evidence reports p95 growth +7.31% / -2.14% / -9.69%, paired 100k wall +1.27%. These gates were not independently rerun in this review. Workflow failures propagate; no hosted-check pass is claimed. | ### New findings **P2: native metadata still depends on inline event boundaries that differ from the editor's visible fragments.** `crates/markdown-view/src/lines.rs:172–190,223–236` resets a fragment at a parsed link label and wrapper end, but not at a subscript/superscript start. It also matches each Text event separately, so a literal character split across Text events can truncate a tag. Fresh probes against the rebuilt native example and installed editor parser give: | Source | Native tags | Editor tags | | --- | --- | --- | | `- 09:00 Meet word~#tag~` | `[]` | `#tag` at [18,22] | | `- 09:00 Meet word^#tag^` | `[]` | `#tag` at [18,22] | | `- 09:00 Meet **#tag_#tag**` | `#tag` at [15,19] | `#tag_` at [15,20] | The first two native titles retain the whole `Meet word~#tag~` / `Meet word^#tag^`; the editor returns `Meet word`. The third case assigns the underscore to different fields. Match the editor's actual fragment boundaries, including opening wrappers and adjacent literal Text events; add independent shared vectors while retaining all six earlier repair vectors. A 675-case ASCII composition sweep found further disagreements in this area; the table gives minimal independently rechecked examples, not a count of distinct defects. **P2: the editor can remove escaped literal content when removing a tag-only wrapper.** At `packages/editor/src/markdown/extraction.ts:73–87,145–158`, `collect` skips non-inline child nodes, then `onlyTags` checks only the remaining visible gaps. An `Escape` can therefore be ignored when deciding that the entire wrapper contains metadata only. For source `- 09:00 Meet *#tag \**` (one literal backslash before the first closing star), both adapters extract `#tag` at [14,18]. The editor returns title `Meet`, removing the escaped literal star too. Native retains title source slices `Meet *` and `\**`. The literal star is not metadata and must survive. Include literal/escaped children when deciding whether a wrapper contains only tags, and add a shared exact-title/source-span regression. ### Cooperative deadline and worst-case time The effective admission caps are source **≤4 MiB**, logical lines **≤16,384**, consecutive blank lines **≤1,024**, and block-container depth **<256**. Node/inline/emphasis counters are each **<250,000**, rows plus relationships **<100,000**, encoded result **<16 MiB**, and the working reservation **<64 MiB**. The reservation models parser memory; it is not a process-wide allocator limit. These caps bound input and counted output/work dimensions. **They do not establish a numeric worst-case elapsed-time bound, including 50 ms.** Admission can scan a long line between checks; normalization, full first-pass parser construction, an individual iterator step, wrapper/tag work and destruction can run outside clock checks. A group of 64 events has no fixed time cost. Offset remapping adds logarithmic endpoint lookup when CRLF is present. The observable return time can exceed the deadline by unchecked work and cleanup, plus scheduling; that excess has no demonstrated numeric maximum here. The 50 ms cooperative deadline is acceptable as a soft refusal mechanism for this unconnected experimental slice. The caps, memory samples and normal-corpus speed pass do not justify a hard 50 ms in-process service guarantee. The current documentation states this limitation. No new P0/P1 was found, so the round-3 stop-and-simplify condition is not triggered. ### Verification All execution was on dev-01, in Nix through `~/bin/slot`, using the existing `calendar-805-r2` worktree and its configured shared target directory. Its commit label is older, but all 13 reviewed runtime/contract/gate files match this head byte-for-byte. Control host: read-only repository/API inspection and review preparation; no builds, tests or installs. Fresh commands (with `CARGO_TARGET_DIR=/home/claude/Developer/Worktrees/calendar-805-r1/target`): ```sh cargo test -j4 -p calternal-markdown-view --locked cargo build -j4 --release -p calternal-markdown-view --example native_gate_probe --locked # packages/editor: bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts ``` The limited Python probes used `resource.setrlimit` before exec: `RLIMIT_AS=(268435456,268435456)`, `RLIMIT_STACK=(8388608,8388608)`, `RLIMIT_CORE=(0,0)`, then `subprocess.run([in_process_containment], input=b"x\r"*2097152, capture_output=True, preexec_fn=caps, timeout=35)`; the second lane adds `--memory`. Native/editor comparisons used the rebuilt `native_gate_probe` and `extract(source, markdownLanguage.parser.parse(normalizeSource(source).slice(bodyOffset(normalizeSource(source)))))`, converting editor UTF-16 ranges to original UTF-8 byte ranges.

kayg referenced this pull request

2026-10-31 21:41:57 +00:00

feat(tasks): port B1 status plans to the Home executor #812

kayg referenced this pull request

2026-10-31 21:56:11 +00:00

feat(tasks): port B1 status plans to the Home executor #812

kayg commented

2026-10-31 22:16:30 +00:00

Author

Owner

Round 3 follow-up at 6514f9fde34ad9653f2a1fce7f1ba5ec30328477.

Native extraction now keeps adjacent text events together and starts fragments at subscript/superscript boundaries. Shared vectors cover both boundary cases and adjacent text events.
Editor tag-only wrapper detection now preserves escaped literal spans. A shared vector covers the escaped-star case.
Grouped the native scanner inputs and collapsed its conditional to clear the two clippy diagnostics.

Fresh dev-01/Nix gates, run through ~/bin/slot:

cargo test -j4 -p calternal-markdown-view --locked: 25 passed
Rust 0067 shared vectors: 1 passed
Targeted task create tests: 12 passed
Editor typecheck: passed
Editor contract tests: 211 passed
cargo fmt -p calternal-markdown-view --check: passed
cargo clippy -j4 -p calternal-markdown-view --all-targets --locked -- -D warnings: passed
Release example build: passed

The requested memory guard and 4/32/256 KiB speed checks remain pending. Both perf hosts had another job holding the required shared lock, so no overlapping measurement ran and no regression result is claimed.

Round 3 follow-up at `6514f9fde34ad9653f2a1fce7f1ba5ec30328477`. - Native extraction now keeps adjacent text events together and starts fragments at subscript/superscript boundaries. Shared vectors cover both boundary cases and adjacent text events. - Editor tag-only wrapper detection now preserves escaped literal spans. A shared vector covers the escaped-star case. - Grouped the native scanner inputs and collapsed its conditional to clear the two clippy diagnostics. Fresh dev-01/Nix gates, run through `~/bin/slot`: - `cargo test -j4 -p calternal-markdown-view --locked`: 25 passed - Rust 0067 shared vectors: 1 passed - Targeted task create tests: 12 passed - Editor typecheck: passed - Editor contract tests: 211 passed - `cargo fmt -p calternal-markdown-view --check`: passed - `cargo clippy -j4 -p calternal-markdown-view --all-targets --locked -- -D warnings`: passed - Release example build: passed The requested memory guard and 4/32/256 KiB speed checks remain pending. Both perf hosts had another job holding the required shared lock, so no overlapping measurement ran and no regression result is claimed.

kayg commented

2026-10-31 23:12:33 +00:00

Author

Owner

Round 3 verification at 4df679942c1a2e5825f920c893d9461eea962ce7. Never merge.

The per-candidate suffix scan now uses a forward cursor for each parser marker (~ and ^). Added the ignored release timing gate lines::scaling_tests::inline_delimiter_search_scales_linearly. It checks five line patterns at 256 KiB, 512 KiB, 1 MiB, 2 MiB and 4 MiB, with a 2.5x ceiling for doublings at the largest sizes.

Scaling

Times are milliseconds for four passes over each line, median of three samples. The largest ratio is 2 MiB to 4 MiB.

Pattern	256 KiB	512 KiB	1 MiB	2 MiB	4 MiB	Largest ratio
`~`	6.454	12.798	25.556	50.867	103.062	2.026x
`~~`	8.243	16.153	32.864	64.935	129.757	1.998x
escaped `*`	3.496	6.973	13.930	28.186	56.202	1.994x
escaped `_`	3.495	7.066	13.983	27.992	56.237	2.009x
backslash runs	3.918	7.846	15.584	30.537	61.813	2.024x

The scaling test passed on perf-02 under perf-lock.

Memory guard

check-markdown-memory.py completed on perf-02 under perf-lock: all admission and full-extraction probes passed, including 72 seeded mixed-ending/blank-run/depth cases. Peak RSS was 30,688 KiB under the 32 MiB guard, with 256 MiB address-space and 8 MiB stack limits.

Paired speed budgets

Three unchanged paired runs used baseline commit 9ce09e32, the original two shards and 100 byte-identical fixtures. One run exceeded the relative p95 growth band at 4 KiB and 32 KiB; its absolute p95 additions were still below 0.2 ms and paired 100k wall growth was +8.96%. The two subsequent runs passed. Median paired deltas across all three runs:

Bucket	Median p95 growth	Median added p95	Gate
4 KiB	+7.35%	+0.003536 ms	pass
32 KiB	+2.52%	+0.008696 ms	pass
256 KiB	-9.87%	-0.325059 ms	pass
100k wall	+2.23%	—	pass (≤10%)

Median direct/head 100k wall was 5.411538 s / 5.532073 s. Each lane handled 925,696,000 bytes, 200,000 rows, 50,000 relationships and zero unresolved notes. The fixture shard SHA-256 values stayed 75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29 and 15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f.

Commands, each under ~/bin/perf-lock on perf-02:

~/bin/perf-lock /home/claude/calendar805-r3/markdown_view_scaling_test --exact lines::scaling_tests::inline_delimiter_search_scales_linearly --ignored --nocapture
~/bin/perf-lock python3 ~/calendar805-r3/check-markdown-memory.py ~/calendar805-r3/in_process_containment
~/bin/perf-lock python3 ~/calendar805-r3/compare-optimized.py ~/calendar805-r3

Fresh dev-01/Nix checks through ~/bin/slot: markdown-view tests (25 passed, timing test ignored by default), Rust shared vectors (1 passed), editor contracts (211 passed), editor typecheck, fmt check and clippy -D warnings all passed. Forgejo currently has no hosted statuses or reviews for this draft head; these are VM results.

Verified release artifact SHA-256: direct baseline 048f981931bb699b744b2863421a6599850f37812cdf45c42e96fcaa68e2d34e; head options-head 28767aca7500cf9d2ccec4c66f5cf1618933d0a021aa7c856a489bab51e19625; in_process_containment 38cef3f187b30db4c10bb93b23a4b7898901773d9fa69776944975e4e2bbd3e3.

Round 3 verification at `4df679942c1a2e5825f920c893d9461eea962ce7`. Never merge. The per-candidate suffix scan now uses a forward cursor for each parser marker (`~` and `^`). Added the ignored release timing gate `lines::scaling_tests::inline_delimiter_search_scales_linearly`. It checks five line patterns at 256 KiB, 512 KiB, 1 MiB, 2 MiB and 4 MiB, with a 2.5x ceiling for doublings at the largest sizes. ### Scaling Times are milliseconds for four passes over each line, median of three samples. The largest ratio is 2 MiB to 4 MiB. | Pattern | 256 KiB | 512 KiB | 1 MiB | 2 MiB | 4 MiB | Largest ratio | | --- | ---: | ---: | ---: | ---: | ---: | ---: | | `~` | 6.454 | 12.798 | 25.556 | 50.867 | 103.062 | 2.026x | | `~~` | 8.243 | 16.153 | 32.864 | 64.935 | 129.757 | 1.998x | | escaped `*` | 3.496 | 6.973 | 13.930 | 28.186 | 56.202 | 1.994x | | escaped `_` | 3.495 | 7.066 | 13.983 | 27.992 | 56.237 | 2.009x | | backslash runs | 3.918 | 7.846 | 15.584 | 30.537 | 61.813 | 2.024x | The scaling test passed on perf-02 under `perf-lock`. ### Memory guard `check-markdown-memory.py` completed on perf-02 under `perf-lock`: all admission and full-extraction probes passed, including 72 seeded mixed-ending/blank-run/depth cases. Peak RSS was 30,688 KiB under the 32 MiB guard, with 256 MiB address-space and 8 MiB stack limits. ### Paired speed budgets Three unchanged paired runs used baseline commit `9ce09e32`, the original two shards and 100 byte-identical fixtures. One run exceeded the relative p95 growth band at 4 KiB and 32 KiB; its absolute p95 additions were still below 0.2 ms and paired 100k wall growth was +8.96%. The two subsequent runs passed. Median paired deltas across all three runs: | Bucket | Median p95 growth | Median added p95 | Gate | | --- | ---: | ---: | --- | | 4 KiB | +7.35% | +0.003536 ms | pass | | 32 KiB | +2.52% | +0.008696 ms | pass | | 256 KiB | -9.87% | -0.325059 ms | pass | | 100k wall | +2.23% | — | pass (≤10%) | Median direct/head 100k wall was 5.411538 s / 5.532073 s. Each lane handled 925,696,000 bytes, 200,000 rows, 50,000 relationships and zero unresolved notes. The fixture shard SHA-256 values stayed `75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29` and `15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f`. Commands, each under `~/bin/perf-lock` on perf-02: ```sh ~/bin/perf-lock /home/claude/calendar805-r3/markdown_view_scaling_test --exact lines::scaling_tests::inline_delimiter_search_scales_linearly --ignored --nocapture ~/bin/perf-lock python3 ~/calendar805-r3/check-markdown-memory.py ~/calendar805-r3/in_process_containment ~/bin/perf-lock python3 ~/calendar805-r3/compare-optimized.py ~/calendar805-r3 ``` Fresh dev-01/Nix checks through `~/bin/slot`: markdown-view tests (25 passed, timing test ignored by default), Rust shared vectors (1 passed), editor contracts (211 passed), editor typecheck, fmt check and clippy `-D warnings` all passed. Forgejo currently has no hosted statuses or reviews for this draft head; these are VM results. Verified release artifact SHA-256: direct baseline `048f981931bb699b744b2863421a6599850f37812cdf45c42e96fcaa68e2d34e`; head `options-head` `28767aca7500cf9d2ccec4c66f5cf1618933d0a021aa7c856a489bab51e19625`; `in_process_containment` `38cef3f187b30db4c10bb93b23a4b7898901773d9fa69776944975e4e2bbd3e3`.

kayg commented

2026-10-31 23:19:39 +00:00

Author

Owner

Independent review: lookahead cache

reviewed 4df679942c1a2e5825f920c893d9461eea962ce7, including #issuecomment-71999.

verdict: cache correctness and scaling pass, with nonblocking P2/P3 follow-ups on performance evidence and coverage. no P0/P1. this is not an unqualified stable-speed-gate pass. no code pushed or merge performed.

findings

P0: none.
P1: none.
P2: speed-gate stability is not established. docs/plans/0067-slice2/compare-optimized.py asserts both <=10% relative p95 growth and <=0.2 ms added p95 for each run, plus <=10% paired wall growth. In the three reports cited by the status (calendar805-r4-perf-final.log, -repeat.log, -confirm.log), 4 KiB relative growth is +15.13%, -5.90%, +7.35%; 32 KiB is +10.0107%, -15.98%, +2.52%. The first run really fails the unchanged relative gate, despite absolute additions of only 5.991 µs / 26.820 µs. Baseline 4 KiB p95 varies 39.604–54.172 µs, while head varies 45.595–51.676 µs. The harness always measures baseline before head, so temporal drift can change the comparison. The reported three-run median is accurate descriptive evidence, but does not replace the script's per-run acceptance rule. Before claiming a stable pass, define a fixed repeated-run acceptance policy in advance and alternate/interleave baseline/head order; retain failed runs and unchanged thresholds. This qualifies measurement confidence, not cache correctness.
P3: scaling coverage omits caret and a fully escaped-closer suffix. Every pattern in crates/markdown-view/src/lines.rs:630 queries b'~'; the asterisk/underscore patterns contain escaped punctuation but still query tilde. Add caret and all-escaped tilde/caret suffix patterns, ideally alongside a small permanent differential test. This does not invalidate the present double-tilde regression gate.

correctness and differential probe

The cache preserves the old scan under the actual caller's invariants: one immutable full source per extraction, forward candidate offsets, and one fixed end per logical line. Each marker has independent state. A cached closer before the next start is discarded; scanning resumes after that closer. A new line starts beyond the preceding line's scanned end. Each extraction creates a fresh cache.

Odd/even backslash parity and adjacent-tilde rejection are unchanged. Backslash runs preceding distinct candidate markers do not overlap, so their backward checks remain amortized linear. Star/underscore formatting and code spans affect parser events, but do not introduce additional cache markers; lexical lookahead through code text is identical to the old scan. Backward requests take the reset path. This private helper is not a general cache for changing source buffers or arbitrary end changes within a line.

On dev-01, an uncommitted detached worktree at ~/Developer/Worktrees/review805-lookahead contains the exact pre-cache function from 6514f9fd and an assertion comparing it with every cached result. Seed 0x8054df67994, 150,000 generated sources, 4,990,569 direct queries, 150,000 production-parser extractions, zero mismatches. Inputs combine single/double/triple tildes, carets, stars, underscores, odd/even backslash runs, escaped markers, code spans, links, tags, Unicode, and LF/CRLF/lone-CR lines. Direct queries retain state across lines, skip candidates to pass closers, alternate markers, and separately exercise backward requests. Production extraction also compares every actual lookahead call. There is no mismatching input to report.

Fresh dev-01/Nix checks through ~/bin/slot:

CARGO_TARGET_DIR=/home/claude/Developer/Worktrees/calendar-805-r1/target cargo test -j4 --release -p calternal-markdown-view --locked seeded_differential -- --nocapture: pass.
Same command without the test filter: 25 existing tests + 1 probe passed, timing test ignored.
In the unchanged calendar805-r4/packages/editor: bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts: 211 passed. Both adapters pass the same authored expectations; all three contracts/markdown-extraction/*.json hashes match between worktrees.

Probe/native/editor logs: dev-01 ~/review805-lookahead-probe.log, ~/review805-native-contract.log, ~/review805-editor-contract.log. Only the probe/instrumentation and required Nix files remain uncommitted.

scaling assessment

yes, the gate is sensitive to the old quadratic scan. In the repeated ~~ pattern, every tilde is rejected as adjacent, so the old function rescans all remaining candidates on every request. For m markers this is m(m−1)/2 candidate visits: doubling size approaches 4x, exceeding the 2.5x threshold. The cache scans that suffix once and subsequent requests are constant work, yielding O(n) total work. The ~ and several other patterns alone would not catch the old behavior because they quickly find a valid closer; the double-tilde case supplies the distinguishing fixture.

The recorded locked perf-02 doubling ratios near 2x support this analysis. I inspected those results and the three raw speed reports, and verified the current baseline/head artifact hashes against #issuecomment-71999. I did not rerun timing measurements on dev-01 or claim an end-to-end worst-case time bound.

## Independent review: lookahead cache reviewed `4df679942c1a2e5825f920c893d9461eea962ce7`, including #issuecomment-71999. **verdict: cache correctness and scaling pass, with nonblocking P2/P3 follow-ups on performance evidence and coverage. no P0/P1.** this is not an unqualified stable-speed-gate pass. no code pushed or merge performed. ### findings - **P0: none.** - **P1: none.** - **P2: speed-gate stability is not established.** `docs/plans/0067-slice2/compare-optimized.py` asserts both <=10% relative p95 growth and <=0.2 ms added p95 for each run, plus <=10% paired wall growth. In the three reports cited by the status (`calendar805-r4-perf-final.log`, `-repeat.log`, `-confirm.log`), 4 KiB relative growth is **+15.13%, -5.90%, +7.35%**; 32 KiB is **+10.0107%, -15.98%, +2.52%**. The first run really fails the unchanged relative gate, despite absolute additions of only **5.991 µs / 26.820 µs**. Baseline 4 KiB p95 varies **39.604–54.172 µs**, while head varies **45.595–51.676 µs**. The harness always measures baseline before head, so temporal drift can change the comparison. The reported three-run median is accurate descriptive evidence, but does not replace the script's per-run acceptance rule. Before claiming a stable pass, define a fixed repeated-run acceptance policy in advance and alternate/interleave baseline/head order; retain failed runs and unchanged thresholds. This qualifies measurement confidence, not cache correctness. - **P3: scaling coverage omits caret and a fully escaped-closer suffix.** Every pattern in `crates/markdown-view/src/lines.rs:630` queries `b'~'`; the asterisk/underscore patterns contain escaped punctuation but still query tilde. Add caret and all-escaped tilde/caret suffix patterns, ideally alongside a small permanent differential test. This does not invalidate the present double-tilde regression gate. ### correctness and differential probe The cache preserves the old scan under the actual caller's invariants: one immutable full source per extraction, forward candidate offsets, and one fixed end per logical line. Each marker has independent state. A cached closer before the next start is discarded; scanning resumes after that closer. A new line starts beyond the preceding line's scanned end. Each extraction creates a fresh cache. Odd/even backslash parity and adjacent-tilde rejection are unchanged. Backslash runs preceding distinct candidate markers do not overlap, so their backward checks remain amortized linear. Star/underscore formatting and code spans affect parser events, but do not introduce additional cache markers; lexical lookahead through code text is identical to the old scan. Backward requests take the reset path. This private helper is not a general cache for changing source buffers or arbitrary end changes within a line. On dev-01, an **uncommitted detached worktree** at `~/Developer/Worktrees/review805-lookahead` contains the exact pre-cache function from `6514f9fd` and an assertion comparing it with every cached result. Seed **`0x8054df67994`**, **150,000 generated sources**, **4,990,569 direct queries**, **150,000 production-parser extractions**, **zero mismatches**. Inputs combine single/double/triple tildes, carets, stars, underscores, odd/even backslash runs, escaped markers, code spans, links, tags, Unicode, and LF/CRLF/lone-CR lines. Direct queries retain state across lines, skip candidates to pass closers, alternate markers, and separately exercise backward requests. Production extraction also compares every actual lookahead call. There is no mismatching input to report. Fresh dev-01/Nix checks through `~/bin/slot`: - `CARGO_TARGET_DIR=/home/claude/Developer/Worktrees/calendar-805-r1/target cargo test -j4 --release -p calternal-markdown-view --locked seeded_differential -- --nocapture`: pass. - Same command without the test filter: **25 existing tests + 1 probe passed**, timing test ignored. - In the unchanged `calendar805-r4/packages/editor`: `bunx --no-install vitest run test/calendar-0067.spike.test.ts test/markdown-contract.test.ts test/markdown-rows.test.ts`: **211 passed**. Both adapters pass the same authored expectations; all three `contracts/markdown-extraction/*.json` hashes match between worktrees. Probe/native/editor logs: dev-01 `~/review805-lookahead-probe.log`, `~/review805-native-contract.log`, `~/review805-editor-contract.log`. Only the probe/instrumentation and required Nix files remain uncommitted. ### scaling assessment **yes, the gate is sensitive to the old quadratic scan.** In the repeated `~~ ` pattern, every tilde is rejected as adjacent, so the old function rescans all remaining candidates on every request. For m markers this is m(m−1)/2 candidate visits: doubling size approaches **4x**, exceeding the **2.5x** threshold. The cache scans that suffix once and subsequent requests are constant work, yielding O(n) total work. The `~ ` and several other patterns alone would not catch the old behavior because they quickly find a valid closer; the double-tilde case supplies the distinguishing fixture. The recorded locked perf-02 doubling ratios near 2x support this analysis. I inspected those results and the three raw speed reports, and verified the current baseline/head artifact hashes against #issuecomment-71999. I did not rerun timing measurements on dev-01 or claim an end-to-end worst-case time bound.

kayg referenced this pull request

2026-11-01 08:22:22 +00:00

Train BM5 #826

kayg referenced this pull request

2026-11-01 08:35:51 +00:00

Train BM5 #826

kayg referenced this pull request

2026-11-01 09:12:32 +00:00

Train BM5 #826

kayg commented

2026-11-02 12:41:55 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/all-build-evidence.txt

Source commit: 515559bbbb8f449d48b4487af98b8b9723842c5a; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 1.51s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c)

running 1 test
test bounded::tests::deadline_is_exclusive ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be)

running 5 tests
test existing_composer_configuration_matches_authored_titles ... ok
test raw_configured_parser_label_evidence ... ok
test authored_link_contract ... ok
test authored_source_rows ... ok
test original_line_contract_through_candidate ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263)

running 5 tests
test literal_markers_do_not_count_as_containers ... ok
test hostile_depth_fails_without_partial_extraction ... ok
test combined_depth_boundaries ... ok
test multiline_list_depth_and_container_pops ... ok
test seeded_mixed_containers_have_one_combined_budget ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32)

running 4 tests
test encoded_result_boundary_counts_all_json_bytes ... ok
test every_count_boundary_is_exclusive ... ok
test combined_seeded_fuzz_preserves_good_generation ... ok
test default_source_and_node_boundaries ... ok

test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.34s

     Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9)

running 2 tests
test text_events_across_lines_keep_parent_and_literal_spans ... ok
test nested_and_mixed_wrappers_preserve_literal_titles ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

    Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view)
    Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.40s
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view)
    Finished `release` profile [optimized] target(s) in 17.10s

ADR 0003 evidence migration: `docs/plans/0067-slice2/all-build-evidence.txt` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 1.51s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c) running 1 test test bounded::tests::deadline_is_exclusive ... ok test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be) running 5 tests test existing_composer_configuration_matches_authored_titles ... ok test raw_configured_parser_label_evidence ... ok test authored_link_contract ... ok test authored_source_rows ... ok test original_line_contract_through_candidate ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263) running 5 tests test literal_markers_do_not_count_as_containers ... ok test hostile_depth_fails_without_partial_extraction ... ok test combined_depth_boundaries ... ok test multiline_list_depth_and_container_pops ... ok test seeded_mixed_containers_have_one_combined_budget ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32) running 4 tests test encoded_result_boundary_counts_all_json_bytes ... ok test every_count_boundary_is_exclusive ... ok test combined_seeded_fuzz_preserves_good_generation ... ok test default_source_and_node_boundaries ... ok test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.34s Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9) running 2 tests test text_events_across_lines_keep_parent_and_literal_spans ... ok test nested_and_mixed_wrappers_preserve_literal_titles ... ok test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view) Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.40s Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view) Finished `release` profile [optimized] target(s) in 17.10s ```

kayg commented

2026-11-02 12:41:56 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/all-perf-evidence.txt

PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=6.288
PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.469
PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.139
PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=1.111
PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.746
PASS depth-256: ('unresolved', 'depth-limit'), bytes=520, elapsed_ms=1.145
PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.182
PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=1.144
PASS mixed-0-depth-256: ('unresolved', 'depth-limit'), bytes=577, elapsed_ms=1.555
PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=3.078
PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.441
PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=2.605
PASS mixed-4-depth-256: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.332
PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=1.101
PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.128
PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=2.999
PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.984
PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.828
PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.677
PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.686
PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=3.323
PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=1.435
PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.240
PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=2.663
PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=2.817
PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.674
PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=1.191
PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=2.722
PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=1.404
PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=3.477
PASS mixed-22-depth-256: ('unresolved', 'depth-limit'), bytes=597, elapsed_ms=1.655
PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.143
PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=3.511
PASS list-depth-256: ('unresolved', 'depth-limit'), bytes=67075, elapsed_ms=1.724
PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.533
PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=16.295
PASS huge-at-limit: ('unresolved', 'source-limit'), bytes=8388608, elapsed_ms=8.851
PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=9.000
PASS formatted-tag-storm-deadline: ('accepted', None), bytes=147464, elapsed_ms=44.560
PASS nodes-below-limit: ('accepted', None), bytes=749993, elapsed_ms=41.983
PASS nodes-at-limit: ('unresolved', 'node-limit'), bytes=749996, elapsed_ms=42.287
PASS nodes-over-limit: ('unresolved', 'node-limit'), bytes=749999, elapsed_ms=41.507
SUMMARY failures=0 names=[]
PASS containment-last-good-queue {"startup_ms": 62.15145200258121, "worker_rss_kib": [5208, 5268], "restarts": 6, "refusals": {"invalid-utf8": "invalid-utf8", "source-limit": "source-limit", "depth-limit": "depth-limit", "worker-death": "worker-death", "timeout": "timeout", "oom": "worker-death/replaced", "stack": "worker-death/replaced"}, "queue": 64, "reserved_foreground": 8, "last_good_restart": true}
MEASURE /home/claude/calendar-805-all/worker-baseline {"run": 0, "notes": 10000, "bytes": 92569600, "unresolved": 0, "rows": 20000, "edges": 5000, "hashes": ["a90857affbbd95b62c6d80233e9c1cf87f86aa74700230fe48269e0879168a2c", "75b786352860955a22ce9ea9e43c3ac3b510a3713ab3b2dff187904b0fccb322"], "startup_ms": 54.90570399706485, "wall_s": 1.2570013280055718, "describe_ms": {"4096": {"50": 0.127018, "95": 0.254327, "99": 0.411581}, "32768": {"50": 0.430096, "95": 0.622547, "99": 0.865752}, "262144": {"50": 3.333096, "95": 4.089324, "99": 4.24202}}, "worker_peak_rss_kib": [5332, 6940], "restarts": 0, "worker_cpu_s": 1.283284}
PASS baseline-10k "recorded"
MEASURE /home/claude/calendar-805-all/worker-head {"run": 0, "notes": 10000, "bytes": 92569600, "unresolved": 0, "rows": 20000, "edges": 5000, "hashes": ["a90857affbbd95b62c6d80233e9c1cf87f86aa74700230fe48269e0879168a2c", "75b786352860955a22ce9ea9e43c3ac3b510a3713ab3b2dff187904b0fccb322"], "startup_ms": 52.84231599944178, "wall_s": 1.8365892240035464, "describe_ms": {"4096": {"50": 0.177112, "95": 0.292949, "99": 0.422984}, "32768": {"50": 0.766356, "95": 0.984043, "99": 1.165905}, "262144": {"50": 5.87545, "95": 6.659461, "99": 7.007492}}, "worker_peak_rss_kib": [5248, 6780], "restarts": 0, "worker_cpu_s": 2.362508}
PASS head-10k "recorded"
FAIL describe-p95-4096 AssertionError({'baseline_ms': 0.254327, 'head_ms': 0.292949, 'budget_ms': 0.2})
FAIL describe-p95-32768 AssertionError({'baseline_ms': 0.622547, 'head_ms': 0.984043, 'budget_ms': 0.2})
FAIL describe-p95-262144 AssertionError({'baseline_ms': 4.089324, 'head_ms': 6.659461, 'budget_ms': 0.2})
MEASURE /home/claude/calendar-805-all/worker-baseline {"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 67.94034000631655, "wall_s": 12.613902855999186, "describe_ms": {"4096": {"50": 0.122529, "95": 0.255859, "99": 0.521207}, "32768": {"50": 0.425356, "95": 0.648195, "99": 1.056089}, "262144": {"50": 3.325763, "95": 4.214518, "99": 5.30918}}, "worker_peak_rss_kib": [5200, 7012], "restarts": 0, "worker_cpu_s": 12.478964}
PASS baseline-100k [{"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 67.94034000631655, "wall_s": 12.613902855999186, "describe_ms": {"4096": {"50": 0.122529, "95": 0.255859, "99": 0.521207}, "32768": {"50": 0.425356, "95": 0.648195, "99": 1.056089}, "262144": {"50": 3.325763, "95": 4.214518, "99": 5.30918}}, "worker_peak_rss_kib": [5200, 7012], "restarts": 0, "worker_cpu_s": 12.478964}]
MEASURE /home/claude/calendar-805-all/worker-head {"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 53.874940997047815, "wall_s": 19.05284061600105, "describe_ms": {"4096": {"50": 0.176782, "95": 0.323006, "99": 0.667641}, "32768": {"50": 0.770865, "95": 1.016575, "99": 1.519968}, "262144": {"50": 5.972492, "95": 6.998465, "99": 8.318689}}, "worker_peak_rss_kib": [5240, 7024], "restarts": 0, "worker_cpu_s": 23.703111}
PASS head-100k [{"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 53.874940997047815, "wall_s": 19.05284061600105, "describe_ms": {"4096": {"50": 0.176782, "95": 0.323006, "99": 0.667641}, "32768": {"50": 0.770865, "95": 1.016575, "99": 1.519968}, "262144": {"50": 5.972492, "95": 6.998465, "99": 8.318689}}, "worker_peak_rss_kib": [5240, 7024], "restarts": 0, "worker_cpu_s": 23.703111}]
PASS incremental-revision-cache {"initial": 1, "unchanged": 0, "edit": 1, "scope": "actual candidate pool revision cache; editor/application not integrated"}
SUMMARY {"failures": ["describe-p95-4096", "describe-p95-32768", "describe-p95-262144"], "repeats": "one each; design/budget failure stops acceptance, no 3-cold/5-warm claim", "unproven": ["application ledger/DB/journal stages", "concurrent durable ack/view latency", "editor changed-region budget", "supervisor 64MiB worst-case decoded-result ceiling", "one retry per revision"]}

ADR 0003 evidence migration: `docs/plans/0067-slice2/all-perf-evidence.txt` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=6.288 PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.469 PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.139 PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=1.111 PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.746 PASS depth-256: ('unresolved', 'depth-limit'), bytes=520, elapsed_ms=1.145 PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.182 PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=1.144 PASS mixed-0-depth-256: ('unresolved', 'depth-limit'), bytes=577, elapsed_ms=1.555 PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=3.078 PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.441 PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=2.605 PASS mixed-4-depth-256: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.332 PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=1.101 PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.128 PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=2.999 PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.984 PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.828 PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.677 PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.686 PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=3.323 PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=1.435 PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.240 PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=2.663 PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=2.817 PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.674 PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=1.191 PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=2.722 PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=1.404 PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=3.477 PASS mixed-22-depth-256: ('unresolved', 'depth-limit'), bytes=597, elapsed_ms=1.655 PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.143 PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=3.511 PASS list-depth-256: ('unresolved', 'depth-limit'), bytes=67075, elapsed_ms=1.724 PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.533 PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=16.295 PASS huge-at-limit: ('unresolved', 'source-limit'), bytes=8388608, elapsed_ms=8.851 PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=9.000 PASS formatted-tag-storm-deadline: ('accepted', None), bytes=147464, elapsed_ms=44.560 PASS nodes-below-limit: ('accepted', None), bytes=749993, elapsed_ms=41.983 PASS nodes-at-limit: ('unresolved', 'node-limit'), bytes=749996, elapsed_ms=42.287 PASS nodes-over-limit: ('unresolved', 'node-limit'), bytes=749999, elapsed_ms=41.507 SUMMARY failures=0 names=[] PASS containment-last-good-queue {"startup_ms": 62.15145200258121, "worker_rss_kib": [5208, 5268], "restarts": 6, "refusals": {"invalid-utf8": "invalid-utf8", "source-limit": "source-limit", "depth-limit": "depth-limit", "worker-death": "worker-death", "timeout": "timeout", "oom": "worker-death/replaced", "stack": "worker-death/replaced"}, "queue": 64, "reserved_foreground": 8, "last_good_restart": true} MEASURE /home/claude/calendar-805-all/worker-baseline {"run": 0, "notes": 10000, "bytes": 92569600, "unresolved": 0, "rows": 20000, "edges": 5000, "hashes": ["a90857affbbd95b62c6d80233e9c1cf87f86aa74700230fe48269e0879168a2c", "75b786352860955a22ce9ea9e43c3ac3b510a3713ab3b2dff187904b0fccb322"], "startup_ms": 54.90570399706485, "wall_s": 1.2570013280055718, "describe_ms": {"4096": {"50": 0.127018, "95": 0.254327, "99": 0.411581}, "32768": {"50": 0.430096, "95": 0.622547, "99": 0.865752}, "262144": {"50": 3.333096, "95": 4.089324, "99": 4.24202}}, "worker_peak_rss_kib": [5332, 6940], "restarts": 0, "worker_cpu_s": 1.283284} PASS baseline-10k "recorded" MEASURE /home/claude/calendar-805-all/worker-head {"run": 0, "notes": 10000, "bytes": 92569600, "unresolved": 0, "rows": 20000, "edges": 5000, "hashes": ["a90857affbbd95b62c6d80233e9c1cf87f86aa74700230fe48269e0879168a2c", "75b786352860955a22ce9ea9e43c3ac3b510a3713ab3b2dff187904b0fccb322"], "startup_ms": 52.84231599944178, "wall_s": 1.8365892240035464, "describe_ms": {"4096": {"50": 0.177112, "95": 0.292949, "99": 0.422984}, "32768": {"50": 0.766356, "95": 0.984043, "99": 1.165905}, "262144": {"50": 5.87545, "95": 6.659461, "99": 7.007492}}, "worker_peak_rss_kib": [5248, 6780], "restarts": 0, "worker_cpu_s": 2.362508} PASS head-10k "recorded" FAIL describe-p95-4096 AssertionError({'baseline_ms': 0.254327, 'head_ms': 0.292949, 'budget_ms': 0.2}) FAIL describe-p95-32768 AssertionError({'baseline_ms': 0.622547, 'head_ms': 0.984043, 'budget_ms': 0.2}) FAIL describe-p95-262144 AssertionError({'baseline_ms': 4.089324, 'head_ms': 6.659461, 'budget_ms': 0.2}) MEASURE /home/claude/calendar-805-all/worker-baseline {"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 67.94034000631655, "wall_s": 12.613902855999186, "describe_ms": {"4096": {"50": 0.122529, "95": 0.255859, "99": 0.521207}, "32768": {"50": 0.425356, "95": 0.648195, "99": 1.056089}, "262144": {"50": 3.325763, "95": 4.214518, "99": 5.30918}}, "worker_peak_rss_kib": [5200, 7012], "restarts": 0, "worker_cpu_s": 12.478964} PASS baseline-100k [{"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 67.94034000631655, "wall_s": 12.613902855999186, "describe_ms": {"4096": {"50": 0.122529, "95": 0.255859, "99": 0.521207}, "32768": {"50": 0.425356, "95": 0.648195, "99": 1.056089}, "262144": {"50": 3.325763, "95": 4.214518, "99": 5.30918}}, "worker_peak_rss_kib": [5200, 7012], "restarts": 0, "worker_cpu_s": 12.478964}] MEASURE /home/claude/calendar-805-all/worker-head {"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 53.874940997047815, "wall_s": 19.05284061600105, "describe_ms": {"4096": {"50": 0.176782, "95": 0.323006, "99": 0.667641}, "32768": {"50": 0.770865, "95": 1.016575, "99": 1.519968}, "262144": {"50": 5.972492, "95": 6.998465, "99": 8.318689}}, "worker_peak_rss_kib": [5240, 7024], "restarts": 0, "worker_cpu_s": 23.703111} PASS head-100k [{"run": 0, "notes": 100000, "bytes": 925696000, "unresolved": 0, "rows": 200000, "edges": 50000, "hashes": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"], "startup_ms": 53.874940997047815, "wall_s": 19.05284061600105, "describe_ms": {"4096": {"50": 0.176782, "95": 0.323006, "99": 0.667641}, "32768": {"50": 0.770865, "95": 1.016575, "99": 1.519968}, "262144": {"50": 5.972492, "95": 6.998465, "99": 8.318689}}, "worker_peak_rss_kib": [5240, 7024], "restarts": 0, "worker_cpu_s": 23.703111}] PASS incremental-revision-cache {"initial": 1, "unchanged": 0, "edit": 1, "scope": "actual candidate pool revision cache; editor/application not integrated"} SUMMARY {"failures": ["describe-p95-4096", "describe-p95-32768", "describe-p95-262144"], "repeats": "one each; design/budget failure stops acceptance, no 3-cold/5-warm claim", "unproven": ["application ledger/DB/journal stages", "concurrent durable ack/view latency", "editor changed-region budget", "supervisor 64MiB worst-case decoded-result ceiling", "one retry per revision"]} ```

kayg commented

2026-11-02 12:41:56 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/depth-green-evidence.txt

Source commit: a85fd82c93e22f676ebab3e091efb6a4adc1e013; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling proc-macro2 v1.0.107
   Compiling quote v1.0.47
   Compiling unicode-ident v1.0.27
   Compiling memchr v2.8.3
   Compiling serde_core v1.0.229
   Compiling aho-corasick v1.1.5
   Compiling serde v1.0.229
   Compiling syn v3.0.9
   Compiling zmij v1.0.23
   Compiling regex-syntax v0.8.11
   Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/vendor/pulldown-cmark)
   Compiling bitflags v2.13.2
   Compiling unicase v2.10.0
   Compiling regex-automata v0.4.18
   Compiling serde_json v1.0.154
   Compiling regex v1.13.1
   Compiling itoa v1.0.18
   Compiling serde_derive v1.0.229
   Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 8.16s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-ba4ab689e69bdd9f)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-10e85e4af17ad063)

running 5 tests
test authored_link_contract ... ok
test existing_composer_configuration_matches_authored_titles ... ok
test authored_source_rows ... ok
test original_line_contract_through_candidate ... ok
test raw_configured_parser_label_evidence ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/depth.rs (target/debug/deps/depth-8ee31a3cd273a9e1)

running 5 tests
test literal_markers_do_not_count_as_containers ... ok
test hostile_depth_fails_without_partial_extraction ... ok
test combined_depth_boundaries ... ok
test seeded_mixed_containers_have_one_combined_budget ... ok
test multiline_list_depth_and_container_pops ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Compiling proc-macro2 v1.0.107
   Compiling quote v1.0.47
   Compiling unicode-ident v1.0.27
   Compiling memchr v2.8.3
   Compiling serde_core v1.0.229
   Compiling aho-corasick v1.1.5
   Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/vendor/pulldown-cmark)
   Compiling regex-syntax v0.8.11
   Compiling syn v3.0.9
   Compiling serde v1.0.229
   Compiling zmij v1.0.23
   Compiling regex-automata v0.4.18
   Compiling serde_derive v1.0.229
   Compiling serde_json v1.0.154
   Compiling unicase v2.10.0
   Compiling bitflags v2.13.2
   Compiling itoa v1.0.18
   Compiling regex v1.13.1
   Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/crates/markdown-view)
    Finished `release` profile [optimized] target(s) in 10.98s
ed3cb30b06121e93a86eae14f3d6f9573fbd4db30d3fe48e1b8e2cd81f74622f  target/release/examples/native_gate_probe
PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=13.648
PASS invalid-utf8: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.018
PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.391
PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=2.233
PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.250
PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=0.881
PASS mixed-0-depth-256: ('accepted', None), bytes=577, elapsed_ms=2.172
PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.883
PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.014
PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=1.416
PASS mixed-4-depth-256: ('accepted', None), bytes=585, elapsed_ms=1.512
PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=0.951
PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.143
PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=1.159
PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.432
PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.254
PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.167
PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.449
PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=1.196
PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=0.851
PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=0.770
PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.348
PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=1.274
PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=1.496
PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=2.061
PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=1.767
PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=2.294
PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=2.436
PASS mixed-22-depth-256: ('accepted', None), bytes=597, elapsed_ms=1.946
PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.137
PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=2.431
PASS list-depth-256: ('accepted', None), bytes=67075, elapsed_ms=2.383
PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.183
PASS huge-at-limit: ('accepted', None), bytes=8388608, elapsed_ms=16.558
PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=15.606

ADR 0003 evidence migration: `docs/plans/0067-slice2/depth-green-evidence.txt` Source commit: [a85fd82c93e22f676ebab3e091efb6a4adc1e013](https://git.kayg.org/kayg/calternal/commit/a85fd82c93e22f676ebab3e091efb6a4adc1e013); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling proc-macro2 v1.0.107 Compiling quote v1.0.47 Compiling unicode-ident v1.0.27 Compiling memchr v2.8.3 Compiling serde_core v1.0.229 Compiling aho-corasick v1.1.5 Compiling serde v1.0.229 Compiling syn v3.0.9 Compiling zmij v1.0.23 Compiling regex-syntax v0.8.11 Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/vendor/pulldown-cmark) Compiling bitflags v2.13.2 Compiling unicase v2.10.0 Compiling regex-automata v0.4.18 Compiling serde_json v1.0.154 Compiling regex v1.13.1 Compiling itoa v1.0.18 Compiling serde_derive v1.0.229 Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 8.16s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-ba4ab689e69bdd9f) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-10e85e4af17ad063) running 5 tests test authored_link_contract ... ok test existing_composer_configuration_matches_authored_titles ... ok test authored_source_rows ... ok test original_line_contract_through_candidate ... ok test raw_configured_parser_label_evidence ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/depth.rs (target/debug/deps/depth-8ee31a3cd273a9e1) running 5 tests test literal_markers_do_not_count_as_containers ... ok test hostile_depth_fails_without_partial_extraction ... ok test combined_depth_boundaries ... ok test seeded_mixed_containers_have_one_combined_budget ... ok test multiline_list_depth_and_container_pops ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Compiling proc-macro2 v1.0.107 Compiling quote v1.0.47 Compiling unicode-ident v1.0.27 Compiling memchr v2.8.3 Compiling serde_core v1.0.229 Compiling aho-corasick v1.1.5 Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/vendor/pulldown-cmark) Compiling regex-syntax v0.8.11 Compiling syn v3.0.9 Compiling serde v1.0.229 Compiling zmij v1.0.23 Compiling regex-automata v0.4.18 Compiling serde_derive v1.0.229 Compiling serde_json v1.0.154 Compiling unicase v2.10.0 Compiling bitflags v2.13.2 Compiling itoa v1.0.18 Compiling regex v1.13.1 Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-s2-805-depth/crates/markdown-view) Finished `release` profile [optimized] target(s) in 10.98s ed3cb30b06121e93a86eae14f3d6f9573fbd4db30d3fe48e1b8e2cd81f74622f target/release/examples/native_gate_probe PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=13.648 PASS invalid-utf8: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.018 PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.391 PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=2.233 PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.250 PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=0.881 PASS mixed-0-depth-256: ('accepted', None), bytes=577, elapsed_ms=2.172 PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.883 PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.014 PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=1.416 PASS mixed-4-depth-256: ('accepted', None), bytes=585, elapsed_ms=1.512 PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=0.951 PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.143 PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=1.159 PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.432 PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.254 PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.167 PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.449 PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=1.196 PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=0.851 PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=0.770 PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.348 PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=1.274 PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=1.496 PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=2.061 PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=1.767 PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=2.294 PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=2.436 PASS mixed-22-depth-256: ('accepted', None), bytes=597, elapsed_ms=1.946 PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.137 PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=2.431 PASS list-depth-256: ('accepted', None), bytes=67075, elapsed_ms=2.383 PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.183 PASS huge-at-limit: ('accepted', None), bytes=8388608, elapsed_ms=16.558 PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=15.606 ```

kayg commented

2026-11-02 12:41:56 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/depth-red-evidence.txt

From ssh://o2.vpn.kayg.org:2222/kayg/calternal
 * branch              feat/calendar-v2-s2-parser-contract -> FETCH_HEAD
   ed324e72..3ae6d32d  feat/calendar-v2-s2-parser-contract -> origin/feat/calendar-v2-s2-parser-contract
HEAD is now at 3ae6d32d test(calendar): cover combined container depth boundaries
evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
    Finished `release` profile [optimized] target(s) in 0.27s
c3cdf2ddcf1a61ddc37f6b263a9fa2fb316caa520ac197d8a5448df80fa032f0  target/release/examples/native_gate_probe
PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=46.923
PASS invalid-utf8: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=6.989
PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=13.545
PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=13.451
FAIL depth-257: expected ('unresolved', 'depth-limit'), got ('accepted', None), rows=1, bytes=522, elapsed_ms=8.410

ADR 0003 evidence migration: `docs/plans/0067-slice2/depth-red-evidence.txt` Source commit: [a85fd82c93e22f676ebab3e091efb6a4adc1e013](https://git.kayg.org/kayg/calternal/commit/a85fd82c93e22f676ebab3e091efb6a4adc1e013); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text From ssh://o2.vpn.kayg.org:2222/kayg/calternal * branch feat/calendar-v2-s2-parser-contract -> FETCH_HEAD ed324e72..3ae6d32d feat/calendar-v2-s2-parser-contract -> origin/feat/calendar-v2-s2-parser-contract HEAD is now at 3ae6d32d test(calendar): cover combined container depth boundaries evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Finished `release` profile [optimized] target(s) in 0.27s c3cdf2ddcf1a61ddc37f6b263a9fa2fb316caa520ac197d8a5448df80fa032f0 target/release/examples/native_gate_probe PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=46.923 PASS invalid-utf8: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=6.989 PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=13.545 PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=13.451 FAIL depth-257: expected ('unresolved', 'depth-limit'), got ('accepted', None), rows=1, bytes=522, elapsed_ms=8.410 ```

kayg commented

2026-11-02 12:41:57 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/evidence.txt

Source commit: 86735e6b7856222d56a7954f1e7543cd04d817f7; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
FORMAT_EXIT=0
    Checking memchr v2.8.3
    Checking serde_core v1.0.229
    Checking bitflags v2.13.2
    Checking unicase v2.10.0
    Checking zmij v1.0.23
    Checking itoa v1.0.18
    Checking pulldown-cmark v0.13.4
    Checking serde v1.0.229
    Checking serde_json v1.0.154
    Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/crates/markdown-view)
    Finished `dev` profile [unoptimized + debuginfo] target(s) in 3.11s
CLIPPY_EXIT=0
    Updating crates.io index
 Downloading crates ...
  Downloaded regex v1.13.1
   Compiling memchr v2.8.3
   Compiling zmij v1.0.23
   Compiling serde_core v1.0.229
   Compiling serde_json v1.0.154
   Compiling regex-syntax v0.8.11
   Compiling aho-corasick v1.1.5
   Compiling pulldown-cmark v0.13.4
   Compiling itoa v1.0.18
   Compiling bitflags v2.13.2
   Compiling unicase v2.10.0
   Compiling regex-automata v0.4.18
   Compiling regex v1.13.1
   Compiling calendar-0067-parity-spike v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/docs/plans/0067-vectors/rust)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 10.29s
     Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef)

running 1 test
70/70 pulldown line vectors passed
test tests::shared_line_vectors ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

   Doc-tests calendar_0067_parity_spike

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

SEED_RUST_EXIT=0
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 1.51s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-fbf4ec72db0b693e)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-f81e52ded74dd611)

running 3 tests
13/13 untouched Composer configuration titles match authored fixtures

thread 'authored_link_contracttest existing_composer_configuration_matches_authored_titles ... ' (ok2784272
) panicked at crates/markdown-view/tests/contract.rs:26:5:
3 / 13 failures:
composer-subscript-compatibility: got {"composer":"Meet H~2~O","links":[{"destination":"p.md","label":"H~2~O","source":"[H~2~O](<p.md>)","span":[5,20]}]}, expected {"composer":"Meet H~2~O","links":[{"destination":"p.md","label":"H2O","source":"[H~2~O](<p.md>)","span":[5,20]}]}
composer-superscript-compatibility: got {"composer":"Meet x^2^","links":[{"destination":"p.md","label":"x^2^","source":"[x^2^](<p.md>)","span":[5,19]}]}, expected {"composer":"Meet x^2^","links":[{"destination":"p.md","label":"x2","source":"[x^2^](<p.md>)","span":[5,19]}]}
composer-strikethrough-compatibility: got {"composer":"Meet old","links":[{"destination":"p.md","label":"old","source":"[~~old~~](<p.md>)","span":[5,22]}]}, expected {"composer":"Meet ~~old~~","links":[{"destination":"p.md","label":"old","source":"[~~old~~](<p.md>)","span":[5,22]}]}
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
test authored_link_contract ... FAILED
[H~2~O](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("H~2~O")), End(Link), End(Paragraph)]
[x^2^](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("x^2")), Text(Borrowed("^")), End(Link), End(Paragraph)]
[~~old~~](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Start(Strikethrough), Text(Borrowed("old")), End(Strikethrough), End(Link), End(Paragraph)]
test raw_configured_parser_label_evidence ... ok

failures:

failures:
    authored_link_contract

test result: FAILED. 2 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

error: test failed, to rerun pass `-p calternal-markdown-view --test contract`
CONTRACT_RUST_EXIT=101

 RUN  v5.0.4 /home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/packages/editor

 ❯ test/markdown-contract.test.ts (13 tests | 3 failed) 58ms
   ❯ shared authored Markdown link contract (13)
     ✓ ordinary-source 11ms
     ✓ composer-person-link 4ms
     ✓ unicode-byte-span 2ms
     ✓ composer-formatting-outside-link 3ms
     ✓ composer-emphasis-in-label 2ms
     ✓ composer-code-in-label 2ms
     ✓ code-does-not-create-link 1ms
     ✓ crlf-byte-span 2ms
     ✓ two-links 2ms
     ✓ emoji-source-label 1ms
     × composer-subscript-compatibility 18ms
     × composer-superscript-compatibility 4ms
     × composer-strikethrough-compatibility 3ms
 ✓ test/calendar-0067.spike.test.ts (70 tests) 74ms

⎯⎯⎯⎯⎯⎯⎯ Failed Tests 3 ⎯⎯⎯⎯⎯⎯⎯

 FAIL  test/markdown-contract.test.ts > shared authored Markdown link contract > composer-subscript-compatibility
AssertionError: expected { links: [ { …(4) } ], …(1) } to deeply equal { links: [ { …(4) } ], …(1) }

- Expected
+ Received

@@ -1,7 +1,7 @@
  {
-   "composer": "Meet H~2~O",
+   "composer": "Meet H2O",
    "links": [
      {
        "destination": "p.md",
        "label": "H2O",
        "source": "[H~2~O](<p.md>)",

 ❯ test/markdown-contract.test.ts:26:17
     24|       composer: view.composer,
     25|     }
     26|     expect(got).toEqual(vector.expected)
       |                 ^
     27|     for (const link of got.links) {
     28|       expect(new TextDecoder('utf-8', { fatal: true }).decode(

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[1/3]⎯

 FAIL  test/markdown-contract.test.ts > shared authored Markdown link contract > composer-superscript-compatibility
AssertionError: expected { links: [ { …(4) } ], …(1) } to deeply equal { links: [ { …(4) } ], …(1) }

- Expected
+ Received

@@ -1,7 +1,7 @@
  {
-   "composer": "Meet x^2^",
+   "composer": "Meet x2",
    "links": [
      {
        "destination": "p.md",
        "label": "x2",
        "source": "[x^2^](<p.md>)",

 ❯ test/markdown-contract.test.ts:26:17
     24|       composer: view.composer,
     25|     }
     26|     expect(got).toEqual(vector.expected)
       |                 ^
     27|     for (const link of got.links) {
     28|       expect(new TextDecoder('utf-8', { fatal: true }).decode(

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[2/3]⎯

 FAIL  test/markdown-contract.test.ts > shared authored Markdown link contract > composer-strikethrough-compatibility
AssertionError: expected { links: [ { …(4) } ], …(1) } to deeply equal { links: [ { …(4) } ], …(1) }

- Expected
+ Received

@@ -1,7 +1,7 @@
  {
-   "composer": "Meet ~~old~~",
+   "composer": "Meet old",
    "links": [
      {
        "destination": "p.md",
        "label": "old",
        "source": "[~~old~~](<p.md>)",

 ❯ test/markdown-contract.test.ts:26:17
     24|       composer: view.composer,
     25|     }
     26|     expect(got).toEqual(vector.expected)
       |                 ^
     27|     for (const link of got.links) {
     28|       expect(new TextDecoder('utf-8', { fatal: true }).decode(

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[3/3]⎯


 Test Files  1 failed | 1 passed (2)
      Tests  3 failed | 80 passed (83)
   Start at  19:29:34
   Duration  1.76s (environment 71%, import 10%, transform 7%, setup 5%, tests 5%, worker 3%)

CONTRACT_TS_EXIT=1
$ tsc --noEmit
TYPECHECK_EXIT=0

ADR 0003 evidence migration: `docs/plans/0067-slice2/evidence.txt` Source commit: [86735e6b7856222d56a7954f1e7543cd04d817f7](https://git.kayg.org/kayg/calternal/commit/86735e6b7856222d56a7954f1e7543cd04d817f7); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead FORMAT_EXIT=0 Checking memchr v2.8.3 Checking serde_core v1.0.229 Checking bitflags v2.13.2 Checking unicase v2.10.0 Checking zmij v1.0.23 Checking itoa v1.0.18 Checking pulldown-cmark v0.13.4 Checking serde v1.0.229 Checking serde_json v1.0.154 Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/crates/markdown-view) Finished `dev` profile [unoptimized + debuginfo] target(s) in 3.11s CLIPPY_EXIT=0 Updating crates.io index Downloading crates ... Downloaded regex v1.13.1 Compiling memchr v2.8.3 Compiling zmij v1.0.23 Compiling serde_core v1.0.229 Compiling serde_json v1.0.154 Compiling regex-syntax v0.8.11 Compiling aho-corasick v1.1.5 Compiling pulldown-cmark v0.13.4 Compiling itoa v1.0.18 Compiling bitflags v2.13.2 Compiling unicase v2.10.0 Compiling regex-automata v0.4.18 Compiling regex v1.13.1 Compiling calendar-0067-parity-spike v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/docs/plans/0067-vectors/rust) Finished `test` profile [unoptimized + debuginfo] target(s) in 10.29s Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef) running 1 test 70/70 pulldown line vectors passed test tests::shared_line_vectors ... ok test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Doc-tests calendar_0067_parity_spike running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s SEED_RUST_EXIT=0 Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 1.51s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-fbf4ec72db0b693e) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-f81e52ded74dd611) running 3 tests 13/13 untouched Composer configuration titles match authored fixtures thread 'authored_link_contracttest existing_composer_configuration_matches_authored_titles ... ' (ok2784272 ) panicked at crates/markdown-view/tests/contract.rs:26:5: 3 / 13 failures: composer-subscript-compatibility: got {"composer":"Meet H~2~O","links":[{"destination":"p.md","label":"H~2~O","source":"[H~2~O](<p.md>)","span":[5,20]}]}, expected {"composer":"Meet H~2~O","links":[{"destination":"p.md","label":"H2O","source":"[H~2~O](<p.md>)","span":[5,20]}]} composer-superscript-compatibility: got {"composer":"Meet x^2^","links":[{"destination":"p.md","label":"x^2^","source":"[x^2^](<p.md>)","span":[5,19]}]}, expected {"composer":"Meet x^2^","links":[{"destination":"p.md","label":"x2","source":"[x^2^](<p.md>)","span":[5,19]}]} composer-strikethrough-compatibility: got {"composer":"Meet old","links":[{"destination":"p.md","label":"old","source":"[~~old~~](<p.md>)","span":[5,22]}]}, expected {"composer":"Meet ~~old~~","links":[{"destination":"p.md","label":"old","source":"[~~old~~](<p.md>)","span":[5,22]}]} note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace test authored_link_contract ... FAILED [H~2~O](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("H~2~O")), End(Link), End(Paragraph)] [x^2^](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("x^2")), Text(Borrowed("^")), End(Link), End(Paragraph)] [~~old~~](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Start(Strikethrough), Text(Borrowed("old")), End(Strikethrough), End(Link), End(Paragraph)] test raw_configured_parser_label_evidence ... ok failures: failures: authored_link_contract test result: FAILED. 2 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s error: test failed, to rerun pass `-p calternal-markdown-view --test contract` CONTRACT_RUST_EXIT=101 RUN v5.0.4 /home/claude/Developer/Worktrees/calendar-v2-s2-parser-contract/packages/editor ❯ test/markdown-contract.test.ts (13 tests | 3 failed) 58ms ❯ shared authored Markdown link contract (13) ✓ ordinary-source 11ms ✓ composer-person-link 4ms ✓ unicode-byte-span 2ms ✓ composer-formatting-outside-link 3ms ✓ composer-emphasis-in-label 2ms ✓ composer-code-in-label 2ms ✓ code-does-not-create-link 1ms ✓ crlf-byte-span 2ms ✓ two-links 2ms ✓ emoji-source-label 1ms × composer-subscript-compatibility 18ms × composer-superscript-compatibility 4ms × composer-strikethrough-compatibility 3ms ✓ test/calendar-0067.spike.test.ts (70 tests) 74ms ⎯⎯⎯⎯⎯⎯⎯ Failed Tests 3 ⎯⎯⎯⎯⎯⎯⎯ FAIL test/markdown-contract.test.ts > shared authored Markdown link contract > composer-subscript-compatibility AssertionError: expected { links: [ { …(4) } ], …(1) } to deeply equal { links: [ { …(4) } ], …(1) } - Expected + Received @@ -1,7 +1,7 @@ { - "composer": "Meet H~2~O", + "composer": "Meet H2O", "links": [ { "destination": "p.md", "label": "H2O", "source": "[H~2~O](<p.md>)", ❯ test/markdown-contract.test.ts:26:17 24| composer: view.composer, 25| } 26| expect(got).toEqual(vector.expected) | ^ 27| for (const link of got.links) { 28| expect(new TextDecoder('utf-8', { fatal: true }).decode( ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[1/3]⎯ FAIL test/markdown-contract.test.ts > shared authored Markdown link contract > composer-superscript-compatibility AssertionError: expected { links: [ { …(4) } ], …(1) } to deeply equal { links: [ { …(4) } ], …(1) } - Expected + Received @@ -1,7 +1,7 @@ { - "composer": "Meet x^2^", + "composer": "Meet x2", "links": [ { "destination": "p.md", "label": "x2", "source": "[x^2^](<p.md>)", ❯ test/markdown-contract.test.ts:26:17 24| composer: view.composer, 25| } 26| expect(got).toEqual(vector.expected) | ^ 27| for (const link of got.links) { 28| expect(new TextDecoder('utf-8', { fatal: true }).decode( ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[2/3]⎯ FAIL test/markdown-contract.test.ts > shared authored Markdown link contract > composer-strikethrough-compatibility AssertionError: expected { links: [ { …(4) } ], …(1) } to deeply equal { links: [ { …(4) } ], …(1) } - Expected + Received @@ -1,7 +1,7 @@ { - "composer": "Meet ~~old~~", + "composer": "Meet old", "links": [ { "destination": "p.md", "label": "old", "source": "[~~old~~](<p.md>)", ❯ test/markdown-contract.test.ts:26:17 24| composer: view.composer, 25| } 26| expect(got).toEqual(vector.expected) | ^ 27| for (const link of got.links) { 28| expect(new TextDecoder('utf-8', { fatal: true }).decode( ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[3/3]⎯ Test Files 1 failed | 1 passed (2) Tests 3 failed | 80 passed (83) Start at 19:29:34 Duration 1.76s (environment 71%, import 10%, transform 7%, setup 5%, tests 5%, worker 3%) CONTRACT_TS_EXIT=1 $ tsc --noEmit TYPECHECK_EXIT=0 ```

kayg commented

2026-11-02 12:41:57 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/final-static-evidence.txt

ADR 0003 evidence migration: `docs/plans/0067-slice2/final-static-evidence.txt` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text ```

kayg commented

2026-11-02 12:41:57 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/linear-build-evidence.txt

Source commit: 62b7304f5280887cb30137591194087cf8a83d5e; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 1.21s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-fdd3608aece89485)

running 5 tests
test existing_composer_configuration_matches_authored_titles ... ok
test authored_link_contract ... ok
test raw_configured_parser_label_evidence ... ok
test original_line_contract_through_candidate ... ok
test authored_source_rows ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/depth.rs (target/debug/deps/depth-323a359bc1680a0c)

running 5 tests
test literal_markers_do_not_count_as_containers ... ok
test hostile_depth_fails_without_partial_extraction ... ok
test combined_depth_boundaries ... ok
test multiline_list_depth_and_container_pops ... ok
test seeded_mixed_containers_have_one_combined_budget ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/tags.rs (target/debug/deps/tags-a14124022c5347ca)

running 2 tests
test text_events_across_lines_keep_parent_and_literal_spans ... ok
test nested_and_mixed_wrappers_preserve_literal_titles ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

    Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view)
    Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.35s
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view)
    Finished `release` profile [optimized] target(s) in 18.26s

ADR 0003 evidence migration: `docs/plans/0067-slice2/linear-build-evidence.txt` Source commit: [62b7304f5280887cb30137591194087cf8a83d5e](https://git.kayg.org/kayg/calternal/commit/62b7304f5280887cb30137591194087cf8a83d5e); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 1.21s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-fdd3608aece89485) running 5 tests test existing_composer_configuration_matches_authored_titles ... ok test authored_link_contract ... ok test raw_configured_parser_label_evidence ... ok test original_line_contract_through_candidate ... ok test authored_source_rows ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/depth.rs (target/debug/deps/depth-323a359bc1680a0c) running 5 tests test literal_markers_do_not_count_as_containers ... ok test hostile_depth_fails_without_partial_extraction ... ok test combined_depth_boundaries ... ok test multiline_list_depth_and_container_pops ... ok test seeded_mixed_containers_have_one_combined_budget ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/tags.rs (target/debug/deps/tags-a14124022c5347ca) running 2 tests test text_events_across_lines_keep_parent_and_literal_spans ... ok test nested_and_mixed_wrappers_preserve_literal_titles ... ok test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view) Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.35s Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-s2-805-linear/crates/markdown-view) Finished `release` profile [optimized] target(s) in 18.26s ```

kayg commented

2026-11-02 12:41:58 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/linear-native-evidence.txt

factor=1 bytes=18440 parser_p50_ms=0.439 parser_p95_ms=0.606 total_p50_ms=8.732 total_p95_ms=10.957 doubling_p95=0.000
factor=2 bytes=36872 parser_p50_ms=0.877 parser_p95_ms=1.386 total_p50_ms=32.218 total_p95_ms=33.912 doubling_p95=3.095
factor=4 bytes=73736 parser_p50_ms=1.721 parser_p95_ms=2.291 total_p50_ms=131.000 total_p95_ms=136.794 doubling_p95=4.034
factor=8 bytes=147464 parser_p50_ms=3.592 parser_p95_ms=4.674 total_p50_ms=541.004 total_p95_ms=549.712 doubling_p95=4.019
factor=1 bytes=18440 parser_p50_ms=0.462 parser_p95_ms=0.624 total_p50_ms=1.089 total_p95_ms=1.486 doubling_p95=0.000
factor=2 bytes=36872 parser_p50_ms=0.782 parser_p95_ms=1.631 total_p50_ms=1.808 total_p95_ms=2.383 doubling_p95=1.604
factor=4 bytes=73736 parser_p50_ms=1.568 parser_p95_ms=2.434 total_p50_ms=3.741 total_p95_ms=3.889 doubling_p95=1.632
factor=8 bytes=147464 parser_p50_ms=3.137 parser_p95_ms=4.112 total_p50_ms=7.418 total_p95_ms=7.640 doubling_p95=1.964
PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=3.090
PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.274
PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.200
PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=1.122
PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.847
PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=2.840
PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.192
PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=1.215
PASS mixed-0-depth-256: ('accepted', None), bytes=577, elapsed_ms=2.721
PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=2.830
PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.228
PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=2.722
PASS mixed-4-depth-256: ('accepted', None), bytes=585, elapsed_ms=2.819
PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=1.342
PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.075
PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=2.648
PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.673
PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=3.086
PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.927
PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.836
PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.774
PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=1.283
PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.078
PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=2.686
PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=2.624
PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.704
PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=1.158
PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=2.683
PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=1.245
PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=2.561
PASS mixed-22-depth-256: ('accepted', None), bytes=597, elapsed_ms=2.645
PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.189
PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=3.467
PASS list-depth-256: ('accepted', None), bytes=67075, elapsed_ms=3.128
PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=2.273
PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=16.869
PASS huge-at-limit: ('accepted', None), bytes=8388608, elapsed_ms=16.985
PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=12.717
PASS formatted-tag-storm-deadline: ('accepted', None), bytes=147464, elapsed_ms=31.408
PASS nodes-at-limit: ('accepted', None), bytes=749996, elapsed_ms=26.831
FAIL nodes-over-limit: expected ('unresolved', 'node-limit'), got ('accepted', None), rows=0, bytes=749999, elapsed_ms=25.864

ADR 0003 evidence migration: `docs/plans/0067-slice2/linear-native-evidence.txt` Source commit: [62b7304f5280887cb30137591194087cf8a83d5e](https://git.kayg.org/kayg/calternal/commit/62b7304f5280887cb30137591194087cf8a83d5e); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text factor=1 bytes=18440 parser_p50_ms=0.439 parser_p95_ms=0.606 total_p50_ms=8.732 total_p95_ms=10.957 doubling_p95=0.000 factor=2 bytes=36872 parser_p50_ms=0.877 parser_p95_ms=1.386 total_p50_ms=32.218 total_p95_ms=33.912 doubling_p95=3.095 factor=4 bytes=73736 parser_p50_ms=1.721 parser_p95_ms=2.291 total_p50_ms=131.000 total_p95_ms=136.794 doubling_p95=4.034 factor=8 bytes=147464 parser_p50_ms=3.592 parser_p95_ms=4.674 total_p50_ms=541.004 total_p95_ms=549.712 doubling_p95=4.019 factor=1 bytes=18440 parser_p50_ms=0.462 parser_p95_ms=0.624 total_p50_ms=1.089 total_p95_ms=1.486 doubling_p95=0.000 factor=2 bytes=36872 parser_p50_ms=0.782 parser_p95_ms=1.631 total_p50_ms=1.808 total_p95_ms=2.383 doubling_p95=1.604 factor=4 bytes=73736 parser_p50_ms=1.568 parser_p95_ms=2.434 total_p50_ms=3.741 total_p95_ms=3.889 doubling_p95=1.632 factor=8 bytes=147464 parser_p50_ms=3.137 parser_p95_ms=4.112 total_p50_ms=7.418 total_p95_ms=7.640 doubling_p95=1.964 PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=3.090 PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.274 PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.200 PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=1.122 PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.847 PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=2.840 PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.192 PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=1.215 PASS mixed-0-depth-256: ('accepted', None), bytes=577, elapsed_ms=2.721 PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=2.830 PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.228 PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=2.722 PASS mixed-4-depth-256: ('accepted', None), bytes=585, elapsed_ms=2.819 PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=1.342 PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.075 PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=2.648 PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=2.673 PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=3.086 PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.927 PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.836 PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=2.774 PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=1.283 PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.078 PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=2.686 PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=2.624 PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=2.704 PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=1.158 PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=2.683 PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=1.245 PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=2.561 PASS mixed-22-depth-256: ('accepted', None), bytes=597, elapsed_ms=2.645 PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.189 PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=3.467 PASS list-depth-256: ('accepted', None), bytes=67075, elapsed_ms=3.128 PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=2.273 PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=16.869 PASS huge-at-limit: ('accepted', None), bytes=8388608, elapsed_ms=16.985 PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=12.717 PASS formatted-tag-storm-deadline: ('accepted', None), bytes=147464, elapsed_ms=31.408 PASS nodes-at-limit: ('accepted', None), bytes=749996, elapsed_ms=26.831 FAIL nodes-over-limit: expected ('unresolved', 'node-limit'), got ('accepted', None), rows=0, bytes=749999, elapsed_ms=25.864 ```

kayg commented

2026-11-02 12:41:58 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/linear-red-evidence.txt

factor=1 bytes=18440 parser_p50_ms=0.431 parser_p95_ms=0.615 total_p50_ms=8.834 total_p95_ms=9.355 doubling_p95=0.000
factor=2 bytes=36872 parser_p50_ms=0.902 parser_p95_ms=1.209 total_p50_ms=32.102 total_p95_ms=36.017 doubling_p95=3.850

thread 'main' (10064) panicked at crates/markdown-view/examples/extraction_scaling.rs:35:13:
doubling exceeds plan's 2.5x budget
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

ADR 0003 evidence migration: `docs/plans/0067-slice2/linear-red-evidence.txt` Source commit: [62b7304f5280887cb30137591194087cf8a83d5e](https://git.kayg.org/kayg/calternal/commit/62b7304f5280887cb30137591194087cf8a83d5e); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text factor=1 bytes=18440 parser_p50_ms=0.431 parser_p95_ms=0.615 total_p50_ms=8.834 total_p95_ms=9.355 doubling_p95=0.000 factor=2 bytes=36872 parser_p50_ms=0.902 parser_p95_ms=1.209 total_p50_ms=32.102 total_p95_ms=36.017 doubling_p95=3.850 thread 'main' (10064) panicked at crates/markdown-view/examples/extraction_scaling.rs:35:13: doubling exceeds plan's 2.5x budget note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace ```

kayg commented

2026-11-02 12:41:58 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/native-limits-evidence.txt

Source commit: a9bddfa1152bb06625c90ef2cc8af2af295db183; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

From ssh://o2.vpn.kayg.org:2222/kayg/calternal
 * branch              feat/calendar-v2-s2-parser-contract -> FETCH_HEAD
   ba902e97..ed324e72  feat/calendar-v2-s2-parser-contract -> origin/feat/calendar-v2-s2-parser-contract
HEAD is now at ed324e72 test(calendar): add native packaging and OS-limited failure probe
evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling proc-macro2 v1.0.107
   Compiling memchr v2.8.3
   Compiling quote v1.0.47
   Compiling unicode-ident v1.0.27
   Compiling serde_core v1.0.229
   Compiling aho-corasick v1.1.5
   Compiling zmij v1.0.23
   Compiling serde v1.0.229
   Compiling syn v3.0.9
   Compiling regex-syntax v0.8.11
   Compiling pulldown-cmark v0.13.4
   Compiling bitflags v2.13.2
   Compiling regex-automata v0.4.18
   Compiling unicase v2.10.0
   Compiling serde_derive v1.0.229
   Compiling serde_json v1.0.154
   Compiling itoa v1.0.18
   Compiling regex v1.13.1
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/crates/markdown-view)
    Finished `release` profile [optimized] target(s) in 25.32s
c3cdf2ddcf1a61ddc37f6b263a9fa2fb316caa520ac197d8a5448df80fa032f0  target/release/examples/native_gate_probe
PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=13.100
PASS invalid-utf8: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=3.148
FAIL depth-257: expected ('unresolved', 'depth-limit'), got ('accepted', None), rows=1, bytes=524, elapsed_ms=5.633
LIMITS_EXIT=1

ADR 0003 evidence migration: `docs/plans/0067-slice2/native-limits-evidence.txt` Source commit: [a9bddfa1152bb06625c90ef2cc8af2af295db183](https://git.kayg.org/kayg/calternal/commit/a9bddfa1152bb06625c90ef2cc8af2af295db183); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text From ssh://o2.vpn.kayg.org:2222/kayg/calternal * branch feat/calendar-v2-s2-parser-contract -> FETCH_HEAD ba902e97..ed324e72 feat/calendar-v2-s2-parser-contract -> origin/feat/calendar-v2-s2-parser-contract HEAD is now at ed324e72 test(calendar): add native packaging and OS-limited failure probe evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling proc-macro2 v1.0.107 Compiling memchr v2.8.3 Compiling quote v1.0.47 Compiling unicode-ident v1.0.27 Compiling serde_core v1.0.229 Compiling aho-corasick v1.1.5 Compiling zmij v1.0.23 Compiling serde v1.0.229 Compiling syn v3.0.9 Compiling regex-syntax v0.8.11 Compiling pulldown-cmark v0.13.4 Compiling bitflags v2.13.2 Compiling regex-automata v0.4.18 Compiling unicase v2.10.0 Compiling serde_derive v1.0.229 Compiling serde_json v1.0.154 Compiling itoa v1.0.18 Compiling regex v1.13.1 Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/crates/markdown-view) Finished `release` profile [optimized] target(s) in 25.32s c3cdf2ddcf1a61ddc37f6b263a9fa2fb316caa520ac197d8a5448df80fa032f0 target/release/examples/native_gate_probe PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=13.100 PASS invalid-utf8: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=3.148 FAIL depth-257: expected ('unresolved', 'depth-limit'), got ('accepted', None), rows=1, bytes=524, elapsed_ms=5.633 LIMITS_EXIT=1 ```

kayg commented

2026-11-02 12:41:59 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/native-remaining-evidence.txt

Source commit: 41642d671d3eb3b8cf7eb4261e20661b7bd00771; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=15.929
PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.068
PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=0.927
PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=0.783
PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=1.788
PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=2.073
PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=0.736
PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=0.907
PASS mixed-0-depth-256: ('accepted', None), bytes=577, elapsed_ms=1.288
PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.329
PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=0.764
PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=1.316
PASS mixed-4-depth-256: ('accepted', None), bytes=585, elapsed_ms=1.166
PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=0.801
PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=0.749
PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=1.053
PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.105
PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=0.849
PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=0.853
PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=0.821
PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=0.798
PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=0.576
PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.081
PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.446
PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=1.246
PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=0.966
PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=0.711
PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=0.839
PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=0.616
PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=1.048
PASS mixed-22-depth-256: ('accepted', None), bytes=597, elapsed_ms=0.950
PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=0.698
PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=1.997
PASS list-depth-256: ('accepted', None), bytes=67075, elapsed_ms=2.160
PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.161
PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=13.861
PASS huge-at-limit: ('accepted', None), bytes=8388608, elapsed_ms=14.473
PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=12.108
FAIL formatted-tag-storm-deadline: no adapter result before 50 ms; child killed/reaped

ADR 0003 evidence migration: `docs/plans/0067-slice2/native-remaining-evidence.txt` Source commit: [41642d671d3eb3b8cf7eb4261e20661b7bd00771](https://git.kayg.org/kayg/calternal/commit/41642d671d3eb3b8cf7eb4261e20661b7bd00771); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=15.929 PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.068 PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=0.927 PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=0.783 PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=1.788 PASS depth-256: ('accepted', None), bytes=520, elapsed_ms=2.073 PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=0.736 PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=0.907 PASS mixed-0-depth-256: ('accepted', None), bytes=577, elapsed_ms=1.288 PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.329 PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=0.764 PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=1.316 PASS mixed-4-depth-256: ('accepted', None), bytes=585, elapsed_ms=1.166 PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=0.801 PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=0.749 PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=1.053 PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.105 PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=0.849 PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=0.853 PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=0.821 PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=0.798 PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=0.576 PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.081 PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.446 PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=1.246 PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=0.966 PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=0.711 PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=0.839 PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=0.616 PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=1.048 PASS mixed-22-depth-256: ('accepted', None), bytes=597, elapsed_ms=0.950 PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=0.698 PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=1.997 PASS list-depth-256: ('accepted', None), bytes=67075, elapsed_ms=2.160 PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.161 PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=13.861 PASS huge-at-limit: ('accepted', None), bytes=8388608, elapsed_ms=14.473 PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=12.108 FAIL formatted-tag-storm-deadline: no adapter result before 50 ms; child killed/reaped ```

kayg commented

2026-11-02 12:41:59 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/optimization-after-profile-evidence.txt

Source commit: 75a1577f5b98909836146e3f3309a23932a6aca2; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

==3767828== Callgrind, a call-graph generating cache profiler
==3767828== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al.
==3767828== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info
==3767828== Command: target/release/examples/parser_options profile 90 100
==3767828== 
==3767828== For interactive control, run 'callgrind_control -h'.
==3767828== 
==3767828== Events    : Ir
==3767828== Collected : 294541785
==3767828== 
==3767828== I   refs:      294,541,785
==3768217== Callgrind, a call-graph generating cache profiler
==3768217== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al.
==3768217== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info
==3768217== Command: target/release/examples/parser_options profile 99 30
==3768217== 
==3768217== For interactive control, run 'callgrind_control -h'.
==3768217== 
==3768217== Events    : Ir
==3768217== Collected : 697639451
==3768217== 
==3768217== I   refs:      697,639,451
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
	LANGUAGE = (unset),
	LC_ALL = (unset),
	LC_CTYPE = (unset),
	LC_NUMERIC = (unset),
	LC_COLLATE = (unset),
	LC_TIME = (unset),
	LC_MESSAGES = (unset),
	LC_MONETARY = (unset),
	LC_ADDRESS = (unset),
	LC_IDENTIFICATION = (unset),
	LC_MEASUREMENT = (unset),
	LC_PAPER = (unset),
	LC_TELEPHONE = (unset),
	LC_NAME = (unset),
	LANG = "en_US.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
--------------------------------------------------------------------------------
Profile data file '/home/claude/805-opt-32.callgrind' (creator: callgrind-3.27.1)
--------------------------------------------------------------------------------
I1 cache: 
D1 cache: 
LL cache: 
Timerange: Basic block 0 - 69651634
Trigger: Program termination
Profiled target:  target/release/examples/parser_options profile 90 100 (PID 3767828, part 1)
Events recorded:  Ir
Events shown:     Ir
Event sort order: Ir
Thresholds:       99
Include dirs:     
User annotated:   
Auto-annotation:  on

--------------------------------------------------------------------------------
Ir                   
--------------------------------------------------------------------------------
294,541,785 (100.0%)  PROGRAM TOTALS

--------------------------------------------------------------------------------
Ir                    file:function
--------------------------------------------------------------------------------
118,736,361 (40.31%)  ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 29,856,206 (10.14%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 26,222,428 ( 8.90%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 16,560,061 ( 5.62%)  ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 13,026,906 ( 4.42%)  ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 11,912,352 ( 4.04%)  ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 10,966,075 ( 3.72%)  ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 10,081,719 ( 3.42%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  7,165,595 ( 2.43%)  ???:__memcpy_avx_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
  6,187,462 ( 2.10%)  ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  5,909,813 ( 2.01%)  ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  5,479,856 ( 1.86%)  ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  4,701,648 ( 1.60%)  ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
  3,951,524 ( 1.34%)  ???:<regex_automata::hybrid::regex::Regex>::try_search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  3,945,060 ( 1.34%)  ???:<regex_automata::meta::strategy::Core as regex_automata::meta::strategy::Strategy>::search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  2,133,928 ( 0.72%)  ???:<pulldown_cmark::firstpass::FirstPass>::scan_paragraph_interrupt [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  1,855,067 ( 0.63%)  ???:<pulldown_cmark::firstpass::FirstPass>::pop [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  1,225,534 ( 0.42%)  ???:<regex_automata::dfa::onepass::InternalBuilder>::shuffle_states [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  1,223,110 ( 0.42%)  ???:pulldown_cmark::firstpass::scan_paragraph_interrupt_no_table [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  1,155,137 ( 0.39%)  ???:<pulldown_cmark::firstpass::FirstPass>::finish_list [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  1,123,120 ( 0.38%)  ???:pulldown_cmark::scanners::scan_code_fence [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    892,711 ( 0.30%)  ???:_int_malloc [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
    856,362 ( 0.29%)  ???:<regex_automata::dfa::onepass::InternalBuilder>::compile_transition [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    693,114 ( 0.24%)  ???:<regex_syntax::utf8::Utf8Sequences as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    674,175 ( 0.23%)  ???:pulldown_cmark::scanners::scan_setext_heading [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    537,050 ( 0.18%)  ???:<regex_automata::nfa::thompson::compiler::Compiler>::c'2 [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    421,776 ( 0.14%)  ???:<std::sys::pal::unix::time::Timespec>::sub_timespec [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    368,004 ( 0.12%)  ???:free [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
    312,051 ( 0.11%)  ???:<regex_automata::nfa::thompson::nfa::Inner>::add [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    306,303 ( 0.10%)  ???:malloc [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
    289,195 ( 0.10%)  ???:<regex_automata::nfa::thompson::compiler::Compiler>::build_many_from_hir::<&regex_syntax::hir::Hir> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    268,559 ( 0.09%)  ???:<serde_json::ser::Compound<&mut calternal_markdown_view::bounded::Count, serde_json::ser::CompactFormatter> as serde_core::ser::SerializeMap>::serialize_key::<str> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    243,505 ( 0.08%)  ???:_int_free_merge_chunk [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
    224,550 ( 0.08%)  ???:regex_automata::meta::strategy::new [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    220,852 ( 0.07%)  ???:<alloc::raw_vec::RawVecInner>::finish_grow [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    200,927 ( 0.07%)  ???:_int_free_create_chunk [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
    200,889 ( 0.07%)  ???:<std::sys::pal::unix::time::Timespec>::now [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    197,940 ( 0.07%)  ???:_dl_relocate_object_no_relro [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/ld-linux-x86-64.so.2]
    189,072 ( 0.06%)  ???:clock_gettime@@GLIBC_2.17 [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
    187,973 ( 0.06%)  ???:<regex_automata::nfa::thompson::builder::Builder>::patch [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    182,918 ( 0.06%)  ???:<regex_automata::nfa::thompson::builder::Builder>::add [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    182,608 ( 0.06%)  ???:<pulldown_cmark::parse::Parser>::handle_inline_pass1 [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    175,083 ( 0.06%)  ???:<regex_automata::nfa::thompson::compiler::Utf8Compiler>::add [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    172,746 ( 0.06%)  ???:<regex_automata::nfa::thompson::map::Utf8BoundedEntry as alloc::vec::spec_from_elem::SpecFromElem>::from_elem::<alloc::alloc::Global> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    164,242 ( 0.06%)  ???:unlink_chunk.isra.0 [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
    154,546 ( 0.05%)  ???:<regex_syntax::hir::interval::IntervalSet<regex_syntax::hir::ClassUnicodeRange>>::canonicalize [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]

perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
	LANGUAGE = (unset),
	LC_ALL = (unset),
	LC_CTYPE = (unset),
	LC_NUMERIC = (unset),
	LC_COLLATE = (unset),
	LC_TIME = (unset),
	LC_MESSAGES = (unset),
	LC_MONETARY = (unset),
	LC_ADDRESS = (unset),
	LC_IDENTIFICATION = (unset),
	LC_MEASUREMENT = (unset),
	LC_PAPER = (unset),
	LC_TELEPHONE = (unset),
	LC_NAME = (unset),
	LANG = "en_US.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
--------------------------------------------------------------------------------
Profile data file '/home/claude/805-opt-256.callgrind' (creator: callgrind-3.27.1)
--------------------------------------------------------------------------------
I1 cache: 
D1 cache: 
LL cache: 
Timerange: Basic block 0 - 166582288
Trigger: Program termination
Profiled target:  target/release/examples/parser_options profile 99 30 (PID 3768217, part 1)
Events recorded:  Ir
Events shown:     Ir
Event sort order: Ir
Thresholds:       99
Include dirs:     
User annotated:   
Auto-annotation:  on

--------------------------------------------------------------------------------
Ir                   
--------------------------------------------------------------------------------
697,639,451 (100.0%)  PROGRAM TOTALS

--------------------------------------------------------------------------------
Ir                    file:function
--------------------------------------------------------------------------------
288,016,900 (41.28%)  ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 73,026,173 (10.47%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 64,063,701 ( 9.18%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 38,792,873 ( 5.56%)  ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 31,707,869 ( 4.55%)  ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 29,206,038 ( 4.19%)  ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 26,864,538 ( 3.85%)  ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 24,663,352 ( 3.54%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 17,314,579 ( 2.48%)  ???:__memcpy_avx_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
 14,816,791 ( 2.12%)  ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 14,344,165 ( 2.06%)  ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 13,165,018 ( 1.89%)  ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
 11,970,907 ( 1.72%)  ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
  9,647,138 ( 1.38%)  ???:<regex_automata::hybrid::regex::Regex>::try_search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  9,645,650 ( 1.38%)  ???:<regex_automata::meta::strategy::Core as regex_automata::meta::strategy::Strategy>::search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  5,233,854 ( 0.75%)  ???:<pulldown_cmark::firstpass::FirstPass>::scan_paragraph_interrupt [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  4,443,726 ( 0.64%)  ???:<pulldown_cmark::firstpass::FirstPass>::pop [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  2,899,244 ( 0.42%)  ???:pulldown_cmark::firstpass::scan_paragraph_interrupt_no_table [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  2,785,009 ( 0.40%)  ???:<pulldown_cmark::firstpass::FirstPass>::finish_list [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  2,754,660 ( 0.39%)  ???:pulldown_cmark::scanners::scan_code_fence [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  1,651,866 ( 0.24%)  ???:pulldown_cmark::scanners::scan_setext_heading [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
  1,225,534 ( 0.18%)  ???:<regex_automata::dfa::onepass::InternalBuilder>::shuffle_states [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    950,832 ( 0.14%)  ???:<std::sys::pal::unix::time::Timespec>::sub_timespec [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    856,362 ( 0.12%)  ???:<regex_automata::dfa::onepass::InternalBuilder>::compile_transition [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]
    693,114 ( 0.10%)  ???:<regex_syntax::utf8::Utf8Sequences as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options]

ADR 0003 evidence migration: `docs/plans/0067-slice2/optimization-after-profile-evidence.txt` Source commit: [75a1577f5b98909836146e3f3309a23932a6aca2](https://git.kayg.org/kayg/calternal/commit/75a1577f5b98909836146e3f3309a23932a6aca2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text ==3767828== Callgrind, a call-graph generating cache profiler ==3767828== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al. ==3767828== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info ==3767828== Command: target/release/examples/parser_options profile 90 100 ==3767828== ==3767828== For interactive control, run 'callgrind_control -h'. ==3767828== ==3767828== Events : Ir ==3767828== Collected : 294541785 ==3767828== ==3767828== I refs: 294,541,785 ==3768217== Callgrind, a call-graph generating cache profiler ==3768217== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al. ==3768217== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info ==3768217== Command: target/release/examples/parser_options profile 99 30 ==3768217== ==3768217== For interactive control, run 'callgrind_control -h'. ==3768217== ==3768217== Events : Ir ==3768217== Collected : 697639451 ==3768217== ==3768217== I refs: 697,639,451 perl: warning: Setting locale failed. perl: warning: Please check that your locale settings: LANGUAGE = (unset), LC_ALL = (unset), LC_CTYPE = (unset), LC_NUMERIC = (unset), LC_COLLATE = (unset), LC_TIME = (unset), LC_MESSAGES = (unset), LC_MONETARY = (unset), LC_ADDRESS = (unset), LC_IDENTIFICATION = (unset), LC_MEASUREMENT = (unset), LC_PAPER = (unset), LC_TELEPHONE = (unset), LC_NAME = (unset), LANG = "en_US.UTF-8" are supported and installed on your system. perl: warning: Falling back to the standard locale ("C"). -------------------------------------------------------------------------------- Profile data file '/home/claude/805-opt-32.callgrind' (creator: callgrind-3.27.1) -------------------------------------------------------------------------------- I1 cache: D1 cache: LL cache: Timerange: Basic block 0 - 69651634 Trigger: Program termination Profiled target: target/release/examples/parser_options profile 90 100 (PID 3767828, part 1) Events recorded: Ir Events shown: Ir Event sort order: Ir Thresholds: 99 Include dirs: User annotated: Auto-annotation: on -------------------------------------------------------------------------------- Ir -------------------------------------------------------------------------------- 294,541,785 (100.0%) PROGRAM TOTALS -------------------------------------------------------------------------------- Ir file:function -------------------------------------------------------------------------------- 118,736,361 (40.31%) ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 29,856,206 (10.14%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 26,222,428 ( 8.90%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 16,560,061 ( 5.62%) ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 13,026,906 ( 4.42%) ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 11,912,352 ( 4.04%) ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 10,966,075 ( 3.72%) ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 10,081,719 ( 3.42%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 7,165,595 ( 2.43%) ???:__memcpy_avx_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 6,187,462 ( 2.10%) ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 5,909,813 ( 2.01%) ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 5,479,856 ( 1.86%) ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 4,701,648 ( 1.60%) ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 3,951,524 ( 1.34%) ???:<regex_automata::hybrid::regex::Regex>::try_search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 3,945,060 ( 1.34%) ???:<regex_automata::meta::strategy::Core as regex_automata::meta::strategy::Strategy>::search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 2,133,928 ( 0.72%) ???:<pulldown_cmark::firstpass::FirstPass>::scan_paragraph_interrupt [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 1,855,067 ( 0.63%) ???:<pulldown_cmark::firstpass::FirstPass>::pop [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 1,225,534 ( 0.42%) ???:<regex_automata::dfa::onepass::InternalBuilder>::shuffle_states [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 1,223,110 ( 0.42%) ???:pulldown_cmark::firstpass::scan_paragraph_interrupt_no_table [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 1,155,137 ( 0.39%) ???:<pulldown_cmark::firstpass::FirstPass>::finish_list [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 1,123,120 ( 0.38%) ???:pulldown_cmark::scanners::scan_code_fence [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 892,711 ( 0.30%) ???:_int_malloc [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 856,362 ( 0.29%) ???:<regex_automata::dfa::onepass::InternalBuilder>::compile_transition [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 693,114 ( 0.24%) ???:<regex_syntax::utf8::Utf8Sequences as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 674,175 ( 0.23%) ???:pulldown_cmark::scanners::scan_setext_heading [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 537,050 ( 0.18%) ???:<regex_automata::nfa::thompson::compiler::Compiler>::c'2 [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 421,776 ( 0.14%) ???:<std::sys::pal::unix::time::Timespec>::sub_timespec [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 368,004 ( 0.12%) ???:free [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 312,051 ( 0.11%) ???:<regex_automata::nfa::thompson::nfa::Inner>::add [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 306,303 ( 0.10%) ???:malloc [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 289,195 ( 0.10%) ???:<regex_automata::nfa::thompson::compiler::Compiler>::build_many_from_hir::<&regex_syntax::hir::Hir> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 268,559 ( 0.09%) ???:<serde_json::ser::Compound<&mut calternal_markdown_view::bounded::Count, serde_json::ser::CompactFormatter> as serde_core::ser::SerializeMap>::serialize_key::<str> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 243,505 ( 0.08%) ???:_int_free_merge_chunk [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 224,550 ( 0.08%) ???:regex_automata::meta::strategy::new [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 220,852 ( 0.07%) ???:<alloc::raw_vec::RawVecInner>::finish_grow [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 200,927 ( 0.07%) ???:_int_free_create_chunk [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 200,889 ( 0.07%) ???:<std::sys::pal::unix::time::Timespec>::now [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 197,940 ( 0.07%) ???:_dl_relocate_object_no_relro [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/ld-linux-x86-64.so.2] 189,072 ( 0.06%) ???:clock_gettime@@GLIBC_2.17 [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 187,973 ( 0.06%) ???:<regex_automata::nfa::thompson::builder::Builder>::patch [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 182,918 ( 0.06%) ???:<regex_automata::nfa::thompson::builder::Builder>::add [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 182,608 ( 0.06%) ???:<pulldown_cmark::parse::Parser>::handle_inline_pass1 [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 175,083 ( 0.06%) ???:<regex_automata::nfa::thompson::compiler::Utf8Compiler>::add [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 172,746 ( 0.06%) ???:<regex_automata::nfa::thompson::map::Utf8BoundedEntry as alloc::vec::spec_from_elem::SpecFromElem>::from_elem::<alloc::alloc::Global> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 164,242 ( 0.06%) ???:unlink_chunk.isra.0 [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 154,546 ( 0.05%) ???:<regex_syntax::hir::interval::IntervalSet<regex_syntax::hir::ClassUnicodeRange>>::canonicalize [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] perl: warning: Setting locale failed. perl: warning: Please check that your locale settings: LANGUAGE = (unset), LC_ALL = (unset), LC_CTYPE = (unset), LC_NUMERIC = (unset), LC_COLLATE = (unset), LC_TIME = (unset), LC_MESSAGES = (unset), LC_MONETARY = (unset), LC_ADDRESS = (unset), LC_IDENTIFICATION = (unset), LC_MEASUREMENT = (unset), LC_PAPER = (unset), LC_TELEPHONE = (unset), LC_NAME = (unset), LANG = "en_US.UTF-8" are supported and installed on your system. perl: warning: Falling back to the standard locale ("C"). -------------------------------------------------------------------------------- Profile data file '/home/claude/805-opt-256.callgrind' (creator: callgrind-3.27.1) -------------------------------------------------------------------------------- I1 cache: D1 cache: LL cache: Timerange: Basic block 0 - 166582288 Trigger: Program termination Profiled target: target/release/examples/parser_options profile 99 30 (PID 3768217, part 1) Events recorded: Ir Events shown: Ir Event sort order: Ir Thresholds: 99 Include dirs: User annotated: Auto-annotation: on -------------------------------------------------------------------------------- Ir -------------------------------------------------------------------------------- 697,639,451 (100.0%) PROGRAM TOTALS -------------------------------------------------------------------------------- Ir file:function -------------------------------------------------------------------------------- 288,016,900 (41.28%) ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 73,026,173 (10.47%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 64,063,701 ( 9.18%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 38,792,873 ( 5.56%) ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 31,707,869 ( 4.55%) ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 29,206,038 ( 4.19%) ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 26,864,538 ( 3.85%) ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 24,663,352 ( 3.54%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 17,314,579 ( 2.48%) ???:__memcpy_avx_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 14,816,791 ( 2.12%) ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 14,344,165 ( 2.06%) ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 13,165,018 ( 1.89%) ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 11,970,907 ( 1.72%) ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 9,647,138 ( 1.38%) ???:<regex_automata::hybrid::regex::Regex>::try_search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 9,645,650 ( 1.38%) ???:<regex_automata::meta::strategy::Core as regex_automata::meta::strategy::Strategy>::search [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 5,233,854 ( 0.75%) ???:<pulldown_cmark::firstpass::FirstPass>::scan_paragraph_interrupt [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 4,443,726 ( 0.64%) ???:<pulldown_cmark::firstpass::FirstPass>::pop [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 2,899,244 ( 0.42%) ???:pulldown_cmark::firstpass::scan_paragraph_interrupt_no_table [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 2,785,009 ( 0.40%) ???:<pulldown_cmark::firstpass::FirstPass>::finish_list [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 2,754,660 ( 0.39%) ???:pulldown_cmark::scanners::scan_code_fence [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 1,651,866 ( 0.24%) ???:pulldown_cmark::scanners::scan_setext_heading [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 1,225,534 ( 0.18%) ???:<regex_automata::dfa::onepass::InternalBuilder>::shuffle_states [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 950,832 ( 0.14%) ???:<std::sys::pal::unix::time::Timespec>::sub_timespec [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 856,362 ( 0.12%) ???:<regex_automata::dfa::onepass::InternalBuilder>::compile_transition [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] 693,114 ( 0.10%) ???:<regex_syntax::utf8::Utf8Sequences as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-optimize/target/release/examples/parser_options] ```

kayg commented

2026-11-02 12:42:00 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/optimization-build-evidence.txt

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling proc-macro2 v1.0.107
   Compiling quote v1.0.47
   Compiling memchr v2.8.3
   Compiling unicode-ident v1.0.27
   Compiling serde_core v1.0.229
   Compiling zmij v1.0.23
   Compiling aho-corasick v1.1.5
   Compiling syn v3.0.9
   Compiling regex-syntax v0.8.11
   Compiling serde v1.0.229
   Compiling serde_json v1.0.154
   Compiling pulldown-cmark v0.13.4 (/home/claude/Developer/Worktrees/calendar-805-optimize/vendor/pulldown-cmark)
   Compiling regex-automata v0.4.18
   Compiling bitflags v2.13.2
   Compiling unicase v2.10.0
   Compiling itoa v1.0.18
   Compiling regex v1.13.1
   Compiling serde_derive v1.0.229
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 20.45s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c)

running 3 tests
test bounded::tests::deadline_is_exclusive ... ok
test in_process::tests::unwind_and_limit_have_no_replacement_payload ... ok
test in_process::tests::hybrid_routes_boundaries_before_parsing ... ok

test result: ok. 3 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be)

running 5 tests
test existing_composer_configuration_matches_authored_titles ... ok
test authored_link_contract ... ok
test raw_configured_parser_label_evidence ... ok
test original_line_contract_through_candidate ... ok
test authored_source_rows ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263)

running 5 tests
test literal_markers_do_not_count_as_containers ... ok
test hostile_depth_fails_without_partial_extraction ... ok
test combined_depth_boundaries ... ok
test multiline_list_depth_and_container_pops ... ok
test seeded_mixed_containers_have_one_combined_budget ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32)

running 4 tests
test every_count_boundary_is_exclusive ... ok
test encoded_result_boundary_counts_all_json_bytes ... ok
test combined_seeded_fuzz_preserves_good_generation ... ok
test default_source_and_node_boundaries ... ok

test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.27s

     Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9)

running 2 tests
test text_events_across_lines_keep_parent_and_literal_spans ... ok
test nested_and_mixed_wrappers_preserve_literal_titles ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

    Checking memchr v2.8.3
    Checking serde_core v1.0.229
    Checking regex-syntax v0.8.11
    Checking zmij v1.0.23
    Checking itoa v1.0.18
    Checking bitflags v2.13.2
    Checking aho-corasick v1.1.5
    Checking unicase v2.10.0
    Checking pulldown-cmark v0.13.4 (/home/claude/Developer/Worktrees/calendar-805-optimize/vendor/pulldown-cmark)
    Checking regex-automata v0.4.18
    Checking serde_json v1.0.154
    Checking serde v1.0.229
    Checking regex v1.13.1
    Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view)
    Finished `dev` profile [unoptimized + debuginfo] target(s) in 3.35s
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view)
    Finished `release` profile [optimized] target(s) in 29.00s

ADR 0003 evidence migration: `docs/plans/0067-slice2/optimization-build-evidence.txt` Source commit: [75a1577f5b98909836146e3f3309a23932a6aca2](https://git.kayg.org/kayg/calternal/commit/75a1577f5b98909836146e3f3309a23932a6aca2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling proc-macro2 v1.0.107 Compiling quote v1.0.47 Compiling memchr v2.8.3 Compiling unicode-ident v1.0.27 Compiling serde_core v1.0.229 Compiling zmij v1.0.23 Compiling aho-corasick v1.1.5 Compiling syn v3.0.9 Compiling regex-syntax v0.8.11 Compiling serde v1.0.229 Compiling serde_json v1.0.154 Compiling pulldown-cmark v0.13.4 (/home/claude/Developer/Worktrees/calendar-805-optimize/vendor/pulldown-cmark) Compiling regex-automata v0.4.18 Compiling bitflags v2.13.2 Compiling unicase v2.10.0 Compiling itoa v1.0.18 Compiling regex v1.13.1 Compiling serde_derive v1.0.229 Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 20.45s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c) running 3 tests test bounded::tests::deadline_is_exclusive ... ok test in_process::tests::unwind_and_limit_have_no_replacement_payload ... ok test in_process::tests::hybrid_routes_boundaries_before_parsing ... ok test result: ok. 3 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be) running 5 tests test existing_composer_configuration_matches_authored_titles ... ok test authored_link_contract ... ok test raw_configured_parser_label_evidence ... ok test original_line_contract_through_candidate ... ok test authored_source_rows ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263) running 5 tests test literal_markers_do_not_count_as_containers ... ok test hostile_depth_fails_without_partial_extraction ... ok test combined_depth_boundaries ... ok test multiline_list_depth_and_container_pops ... ok test seeded_mixed_containers_have_one_combined_budget ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32) running 4 tests test every_count_boundary_is_exclusive ... ok test encoded_result_boundary_counts_all_json_bytes ... ok test combined_seeded_fuzz_preserves_good_generation ... ok test default_source_and_node_boundaries ... ok test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.27s Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9) running 2 tests test text_events_across_lines_keep_parent_and_literal_spans ... ok test nested_and_mixed_wrappers_preserve_literal_titles ... ok test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Checking memchr v2.8.3 Checking serde_core v1.0.229 Checking regex-syntax v0.8.11 Checking zmij v1.0.23 Checking itoa v1.0.18 Checking bitflags v2.13.2 Checking aho-corasick v1.1.5 Checking unicase v2.10.0 Checking pulldown-cmark v0.13.4 (/home/claude/Developer/Worktrees/calendar-805-optimize/vendor/pulldown-cmark) Checking regex-automata v0.4.18 Checking serde_json v1.0.154 Checking serde v1.0.229 Checking regex v1.13.1 Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view) Finished `dev` profile [unoptimized + debuginfo] target(s) in 3.35s Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view) Finished `release` profile [optimized] target(s) in 29.00s ```

kayg commented

2026-11-02 12:42:00 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/optimization-final-evidence.txt

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 0.78s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c)

running 3 tests
test bounded::tests::deadline_is_exclusive ... ok
test in_process::tests::unwind_and_limit_have_no_replacement_payload ... ok
test in_process::tests::hybrid_routes_boundaries_before_parsing ... ok

test result: ok. 3 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be)

running 5 tests
test existing_composer_configuration_matches_authored_titles ... ok
test authored_link_contract ... ok
test raw_configured_parser_label_evidence ... ok
test original_line_contract_through_candidate ... ok
test authored_source_rows ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263)

running 5 tests
test combined_depth_boundaries ... ok
test literal_markers_do_not_count_as_containers ... ok
test hostile_depth_fails_without_partial_extraction ... ok
test multiline_list_depth_and_container_pops ... ok
test seeded_mixed_containers_have_one_combined_budget ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32)

running 5 tests
test encoded_result_boundary_counts_all_json_bytes ... ok
test every_count_boundary_is_exclusive ... ok
test combined_seeded_fuzz_preserves_good_generation ... ok
test encoded_result_boundary_across_clock_batches ... ok
test default_source_and_node_boundaries ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.32s

     Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9)

running 2 tests
test text_events_across_lines_keep_parent_and_literal_spans ... ok
test nested_and_mixed_wrappers_preserve_literal_titles ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

    Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view)
    Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.24s

ADR 0003 evidence migration: `docs/plans/0067-slice2/optimization-final-evidence.txt` Source commit: [75a1577f5b98909836146e3f3309a23932a6aca2](https://git.kayg.org/kayg/calternal/commit/75a1577f5b98909836146e3f3309a23932a6aca2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 0.78s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c) running 3 tests test bounded::tests::deadline_is_exclusive ... ok test in_process::tests::unwind_and_limit_have_no_replacement_payload ... ok test in_process::tests::hybrid_routes_boundaries_before_parsing ... ok test result: ok. 3 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be) running 5 tests test existing_composer_configuration_matches_authored_titles ... ok test authored_link_contract ... ok test raw_configured_parser_label_evidence ... ok test original_line_contract_through_candidate ... ok test authored_source_rows ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263) running 5 tests test combined_depth_boundaries ... ok test literal_markers_do_not_count_as_containers ... ok test hostile_depth_fails_without_partial_extraction ... ok test multiline_list_depth_and_container_pops ... ok test seeded_mixed_containers_have_one_combined_budget ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32) running 5 tests test encoded_result_boundary_counts_all_json_bytes ... ok test every_count_boundary_is_exclusive ... ok test combined_seeded_fuzz_preserves_good_generation ... ok test encoded_result_boundary_across_clock_batches ... ok test default_source_and_node_boundaries ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.32s Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9) running 2 tests test text_events_across_lines_keep_parent_and_literal_spans ... ok test nested_and_mixed_wrappers_preserve_literal_titles ... ok test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/crates/markdown-view) Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.24s ```

kayg commented

2026-11-02 12:42:00 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/optimization-native-evidence.txt

PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=3.148
PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.525
PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.185
PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=1.238
PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.131
PASS depth-256: ('unresolved', 'depth-limit'), bytes=520, elapsed_ms=1.238
PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.073
PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=1.064
PASS mixed-0-depth-256: ('unresolved', 'depth-limit'), bytes=577, elapsed_ms=1.090
PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.599
PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.482
PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=1.752
PASS mixed-4-depth-256: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.272
PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=1.492
PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.499
PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=1.606
PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.494
PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.675
PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=1.596
PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=1.690
PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=1.513
PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=1.527
PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.693
PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.973
PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=1.814
PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=1.723
PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=1.305
PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=1.811
PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=1.445
PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=2.227
PASS mixed-22-depth-256: ('unresolved', 'depth-limit'), bytes=597, elapsed_ms=1.364
PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.695
PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=2.542
PASS list-depth-256: ('unresolved', 'depth-limit'), bytes=67075, elapsed_ms=1.755
PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.753
PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=16.512
PASS huge-at-limit: ('unresolved', 'source-limit'), bytes=8388608, elapsed_ms=8.884
PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=9.471
PASS formatted-tag-storm-deadline: ('accepted', None), bytes=147464, elapsed_ms=29.367
PASS nodes-below-limit: ('accepted', None), bytes=749993, elapsed_ms=28.667
PASS nodes-at-limit: ('unresolved', 'node-limit'), bytes=749996, elapsed_ms=35.425
PASS nodes-over-limit: ('unresolved', 'node-limit'), bytes=749999, elapsed_ms=28.063
SUMMARY failures=0 names=[]

ADR 0003 evidence migration: `docs/plans/0067-slice2/optimization-native-evidence.txt` Source commit: [75a1577f5b98909836146e3f3309a23932a6aca2](https://git.kayg.org/kayg/calternal/commit/75a1577f5b98909836146e3f3309a23932a6aca2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text PASS ordinary: ('accepted', None), bytes=27, elapsed_ms=3.148 PASS invalid-utf8-lead: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.525 PASS invalid-utf8-continuation: ('unresolved', 'invalid-utf8'), bytes=9, elapsed_ms=1.185 PASS invalid-utf8-truncated: ('unresolved', 'invalid-utf8'), bytes=10, elapsed_ms=1.238 PASS depth-255: ('accepted', None), bytes=518, elapsed_ms=2.131 PASS depth-256: ('unresolved', 'depth-limit'), bytes=520, elapsed_ms=1.238 PASS depth-257: ('unresolved', 'depth-limit'), bytes=522, elapsed_ms=1.073 PASS original-depth-repro: ('unresolved', 'depth-limit'), bytes=524, elapsed_ms=1.064 PASS mixed-0-depth-256: ('unresolved', 'depth-limit'), bytes=577, elapsed_ms=1.090 PASS mixed-1-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.599 PASS mixed-2-depth-257: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.482 PASS mixed-3-depth-255: ('accepted', None), bytes=579, elapsed_ms=1.752 PASS mixed-4-depth-256: ('unresolved', 'depth-limit'), bytes=585, elapsed_ms=1.272 PASS mixed-5-depth-257: ('unresolved', 'depth-limit'), bytes=593, elapsed_ms=1.492 PASS mixed-6-depth-257: ('unresolved', 'depth-limit'), bytes=581, elapsed_ms=1.499 PASS mixed-7-depth-255: ('accepted', None), bytes=581, elapsed_ms=1.606 PASS mixed-8-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.494 PASS mixed-9-depth-255: ('accepted', None), bytes=584, elapsed_ms=1.675 PASS mixed-10-depth-255: ('accepted', None), bytes=585, elapsed_ms=1.596 PASS mixed-11-depth-255: ('accepted', None), bytes=582, elapsed_ms=1.690 PASS mixed-12-depth-255: ('accepted', None), bytes=582, elapsed_ms=1.513 PASS mixed-13-depth-257: ('unresolved', 'depth-limit'), bytes=589, elapsed_ms=1.527 PASS mixed-14-depth-257: ('unresolved', 'depth-limit'), bytes=584, elapsed_ms=1.693 PASS mixed-15-depth-255: ('accepted', None), bytes=580, elapsed_ms=1.973 PASS mixed-16-depth-255: ('accepted', None), bytes=588, elapsed_ms=1.814 PASS mixed-17-depth-255: ('accepted', None), bytes=585, elapsed_ms=1.723 PASS mixed-18-depth-257: ('unresolved', 'depth-limit'), bytes=587, elapsed_ms=1.305 PASS mixed-19-depth-255: ('accepted', None), bytes=578, elapsed_ms=1.811 PASS mixed-20-depth-257: ('unresolved', 'depth-limit'), bytes=592, elapsed_ms=1.445 PASS mixed-21-depth-255: ('accepted', None), bytes=595, elapsed_ms=2.227 PASS mixed-22-depth-256: ('unresolved', 'depth-limit'), bytes=597, elapsed_ms=1.364 PASS mixed-23-depth-257: ('unresolved', 'depth-limit'), bytes=583, elapsed_ms=1.695 PASS list-depth-255: ('accepted', None), bytes=66558, elapsed_ms=2.542 PASS list-depth-256: ('unresolved', 'depth-limit'), bytes=67075, elapsed_ms=1.755 PASS list-depth-257: ('unresolved', 'depth-limit'), bytes=67594, elapsed_ms=1.753 PASS huge-below-limit: ('accepted', None), bytes=8388607, elapsed_ms=16.512 PASS huge-at-limit: ('unresolved', 'source-limit'), bytes=8388608, elapsed_ms=8.884 PASS huge-over-limit: ('unresolved', 'source-limit'), bytes=8388609, elapsed_ms=9.471 PASS formatted-tag-storm-deadline: ('accepted', None), bytes=147464, elapsed_ms=29.367 PASS nodes-below-limit: ('accepted', None), bytes=749993, elapsed_ms=28.667 PASS nodes-at-limit: ('unresolved', 'node-limit'), bytes=749996, elapsed_ms=35.425 PASS nodes-over-limit: ('unresolved', 'node-limit'), bytes=749999, elapsed_ms=28.063 SUMMARY failures=0 names=[] ```

kayg commented

2026-11-02 12:42:01 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/optimization-perf-evidence.txt

FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.035065,"p95_ms":0.051016,"p99_ms":0.062466999999999995},{"bytes":32768,"p50_ms":0.273984,"p95_ms":0.361757,"p99_ms":0.398487},{"bytes":262144,"p50_ms":2.793775,"p95_ms":3.6043540000000003,"p99_ms":3.685777}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.59564179}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.035065, "p95_ms": 0.051016, "p99_ms": 0.062466999999999995}, {"bytes": 32768, "p50_ms": 0.273984, "p95_ms": 0.361757, "p99_ms": 0.398487}, {"bytes": 262144, "p50_ms": 2.793775, "p95_ms": 3.6043540000000003, "p99_ms": 3.685777}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.59564179, "cpu_s": 0.914411, "maxrss_kib": 20036}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.044623,"p95_ms":0.058209,"p99_ms":0.086923},{"bytes":32768,"p50_ms":0.306825,"p95_ms":0.384691,"p99_ms":0.66672},{"bytes":262144,"p50_ms":2.725488,"p95_ms":3.07361,"p99_ms":3.28701}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.59348085}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.044623, "p95_ms": 0.058209, "p99_ms": 0.086923}, {"bytes": 32768, "p50_ms": 0.306825, "p95_ms": 0.384691, "p99_ms": 0.66672}, {"bytes": 262144, "p50_ms": 2.725488, "p95_ms": 3.07361, "p99_ms": 3.28701}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.59348085, "cpu_s": 0.98214, "maxrss_kib": 20036}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.034815,"p95_ms":0.049042,"p99_ms":0.05855},{"bytes":32768,"p50_ms":0.26815100000000003,"p95_ms":0.349985,"p99_ms":0.38484199999999996},{"bytes":262144,"p50_ms":2.656357,"p95_ms":3.242516,"p99_ms":3.6837530000000003}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.584683746}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.034815, "p95_ms": 0.049042, "p99_ms": 0.05855}, {"bytes": 32768, "p50_ms": 0.26815100000000003, "p95_ms": 0.349985, "p99_ms": 0.38484199999999996}, {"bytes": 262144, "p50_ms": 2.656357, "p95_ms": 3.242516, "p99_ms": 3.6837530000000003}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.584683746, "cpu_s": 8.692668, "maxrss_kib": 20036}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.036628999999999995,"p95_ms":0.049221999999999995,"p99_ms":0.058329},{"bytes":32768,"p50_ms":0.25055,"p95_ms":0.316633,"p99_ms":0.35934299999999997},{"bytes":262144,"p50_ms":2.28962,"p95_ms":2.817641,"p99_ms":3.079952}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.175044637}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.036628999999999995, "p95_ms": 0.049221999999999995, "p99_ms": 0.058329}, {"bytes": 32768, "p50_ms": 0.25055, "p95_ms": 0.316633, "p99_ms": 0.35934299999999997}, {"bytes": 262144, "p50_ms": 2.28962, "p95_ms": 2.817641, "p99_ms": 3.079952}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.175044637, "cpu_s": 8.24885, "maxrss_kib": 20036}
BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.00017999999999999267, "growth": 0.00367032339627249, "added_pass": true, "growth_pass": true}, {"bytes": 32768, "added_p95_ms": -0.03335199999999999, "growth": -0.0952955126648285, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.4248750000000001, "growth": -0.13103250685578738, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5897380183129857, "vs_paired_direct": -0.07335045772169313, "historical_wall_pass": true, "paired_wall_pass": true}
CONTAINMENT {"probe": "ordinary", "bytes": 24, "exit": 0, "stdout": "{\"elapsed_ms\":0.532377,\"outcome\":{\"rows\":2,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "depth255", "bytes": 518, "exit": 0, "stdout": "{\"elapsed_ms\":0.513753,\"outcome\":{\"rows\":1,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "depth256", "bytes": 520, "exit": 0, "stdout": "{\"elapsed_ms\":0.036829,\"outcome\":{\"reason\":\"depth-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "long-paragraph", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":5.16036,\"outcome\":{\"rows\":0,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "short-line-storm", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":29.064267,\"outcome\":{\"reason\":\"line-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "inline-storm", "bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":46.386458999999995,\"outcome\":{\"reason\":\"node-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "source-limit", "bytes": 8388608, "exit": 0, "stdout": "{\"elapsed_ms\":0.576771,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}

ADR 0003 evidence migration: `docs/plans/0067-slice2/optimization-perf-evidence.txt` Source commit: [75a1577f5b98909836146e3f3309a23932a6aca2](https://git.kayg.org/kayg/calternal/commit/75a1577f5b98909836146e3f3309a23932a6aca2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.035065,"p95_ms":0.051016,"p99_ms":0.062466999999999995},{"bytes":32768,"p50_ms":0.273984,"p95_ms":0.361757,"p99_ms":0.398487},{"bytes":262144,"p50_ms":2.793775,"p95_ms":3.6043540000000003,"p99_ms":3.685777}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.59564179} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.035065, "p95_ms": 0.051016, "p99_ms": 0.062466999999999995}, {"bytes": 32768, "p50_ms": 0.273984, "p95_ms": 0.361757, "p99_ms": 0.398487}, {"bytes": 262144, "p50_ms": 2.793775, "p95_ms": 3.6043540000000003, "p99_ms": 3.685777}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.59564179, "cpu_s": 0.914411, "maxrss_kib": 20036} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.044623,"p95_ms":0.058209,"p99_ms":0.086923},{"bytes":32768,"p50_ms":0.306825,"p95_ms":0.384691,"p99_ms":0.66672},{"bytes":262144,"p50_ms":2.725488,"p95_ms":3.07361,"p99_ms":3.28701}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.59348085} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.044623, "p95_ms": 0.058209, "p99_ms": 0.086923}, {"bytes": 32768, "p50_ms": 0.306825, "p95_ms": 0.384691, "p99_ms": 0.66672}, {"bytes": 262144, "p50_ms": 2.725488, "p95_ms": 3.07361, "p99_ms": 3.28701}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.59348085, "cpu_s": 0.98214, "maxrss_kib": 20036} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.034815,"p95_ms":0.049042,"p99_ms":0.05855},{"bytes":32768,"p50_ms":0.26815100000000003,"p95_ms":0.349985,"p99_ms":0.38484199999999996},{"bytes":262144,"p50_ms":2.656357,"p95_ms":3.242516,"p99_ms":3.6837530000000003}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.584683746} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.034815, "p95_ms": 0.049042, "p99_ms": 0.05855}, {"bytes": 32768, "p50_ms": 0.26815100000000003, "p95_ms": 0.349985, "p99_ms": 0.38484199999999996}, {"bytes": 262144, "p50_ms": 2.656357, "p95_ms": 3.242516, "p99_ms": 3.6837530000000003}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.584683746, "cpu_s": 8.692668, "maxrss_kib": 20036} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.036628999999999995,"p95_ms":0.049221999999999995,"p99_ms":0.058329},{"bytes":32768,"p50_ms":0.25055,"p95_ms":0.316633,"p99_ms":0.35934299999999997},{"bytes":262144,"p50_ms":2.28962,"p95_ms":2.817641,"p99_ms":3.079952}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.175044637} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.036628999999999995, "p95_ms": 0.049221999999999995, "p99_ms": 0.058329}, {"bytes": 32768, "p50_ms": 0.25055, "p95_ms": 0.316633, "p99_ms": 0.35934299999999997}, {"bytes": 262144, "p50_ms": 2.28962, "p95_ms": 2.817641, "p99_ms": 3.079952}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.175044637, "cpu_s": 8.24885, "maxrss_kib": 20036} BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.00017999999999999267, "growth": 0.00367032339627249, "added_pass": true, "growth_pass": true}, {"bytes": 32768, "added_p95_ms": -0.03335199999999999, "growth": -0.0952955126648285, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.4248750000000001, "growth": -0.13103250685578738, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5897380183129857, "vs_paired_direct": -0.07335045772169313, "historical_wall_pass": true, "paired_wall_pass": true} CONTAINMENT {"probe": "ordinary", "bytes": 24, "exit": 0, "stdout": "{\"elapsed_ms\":0.532377,\"outcome\":{\"rows\":2,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "depth255", "bytes": 518, "exit": 0, "stdout": "{\"elapsed_ms\":0.513753,\"outcome\":{\"rows\":1,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "depth256", "bytes": 520, "exit": 0, "stdout": "{\"elapsed_ms\":0.036829,\"outcome\":{\"reason\":\"depth-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "long-paragraph", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":5.16036,\"outcome\":{\"rows\":0,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "short-line-storm", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":29.064267,\"outcome\":{\"reason\":\"line-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "inline-storm", "bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":46.386458999999995,\"outcome\":{\"reason\":\"node-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "source-limit", "bytes": 8388608, "exit": 0, "stdout": "{\"elapsed_ms\":0.576771,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} ```

kayg commented

2026-11-02 12:42:01 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/optimization-profile-evidence.txt

Source commit: 44165b952c736eb39928c8c3eeb05a42d97b71e8; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

Original file (35,640 bytes): optimization-profile-evidence.txt

ADR 0003 evidence migration: `docs/plans/0067-slice2/optimization-profile-evidence.txt` Source commit: [44165b952c736eb39928c8c3eeb05a42d97b71e8](https://git.kayg.org/kayg/calternal/commit/44165b952c736eb39928c8c3eeb05a42d97b71e8); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. Original file (35,640 bytes): [optimization-profile-evidence.txt](https://git.kayg.org/attachments/344a954f-6a9c-4eaf-a0a1-08cec64bd543)

kayg commented

2026-11-02 12:42:01 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/optimization-vectors-evidence.txt

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling memchr v2.8.3
   Compiling serde_core v1.0.229
   Compiling zmij v1.0.23
   Compiling regex-syntax v0.8.11
   Compiling aho-corasick v1.1.5
   Compiling pulldown-cmark v0.13.4
   Compiling serde_json v1.0.154
   Compiling itoa v1.0.18
   Compiling unicase v2.10.0
   Compiling regex-automata v0.4.18
   Compiling bitflags v2.13.2
   Compiling regex v1.13.1
   Compiling calendar-0067-parity-spike v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/docs/plans/0067-vectors/rust)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 4.80s
     Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef)

running 1 test
test tests::shared_line_vectors ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calendar_0067_parity_spike

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

ADR 0003 evidence migration: `docs/plans/0067-slice2/optimization-vectors-evidence.txt` Source commit: [75a1577f5b98909836146e3f3309a23932a6aca2](https://git.kayg.org/kayg/calternal/commit/75a1577f5b98909836146e3f3309a23932a6aca2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling memchr v2.8.3 Compiling serde_core v1.0.229 Compiling zmij v1.0.23 Compiling regex-syntax v0.8.11 Compiling aho-corasick v1.1.5 Compiling pulldown-cmark v0.13.4 Compiling serde_json v1.0.154 Compiling itoa v1.0.18 Compiling unicase v2.10.0 Compiling regex-automata v0.4.18 Compiling bitflags v2.13.2 Compiling regex v1.13.1 Compiling calendar-0067-parity-spike v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-optimize/docs/plans/0067-vectors/rust) Finished `test` profile [unoptimized + debuginfo] target(s) in 4.80s Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef) running 1 test test tests::shared_line_vectors ... ok test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calendar_0067_parity_spike running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s ```

kayg commented

2026-11-02 12:42:02 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/options-build-evidence.txt

Source commit: 9064e29ac5ca9744cd369c1137a16bfd764e37d3; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-options/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 0.63s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c)

running 3 tests
test bounded::tests::deadline_is_exclusive ... ok
test in_process::tests::hybrid_routes_boundaries_before_parsing ... ok
test in_process::tests::unwind_and_limit_have_no_replacement_payload ... ok

test result: ok. 3 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be)

running 5 tests
test existing_composer_configuration_matches_authored_titles ... ok
test authored_link_contract ... ok
test raw_configured_parser_label_evidence ... ok
test original_line_contract_through_candidate ... ok
test authored_source_rows ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263)

running 5 tests
test literal_markers_do_not_count_as_containers ... ok
test hostile_depth_fails_without_partial_extraction ... ok
test combined_depth_boundaries ... ok
test multiline_list_depth_and_container_pops ... ok
test seeded_mixed_containers_have_one_combined_budget ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32)

running 4 tests
test encoded_result_boundary_counts_all_json_bytes ... ok
test every_count_boundary_is_exclusive ... ok
test combined_seeded_fuzz_preserves_good_generation ... ok
test default_source_and_node_boundaries ... ok

test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.35s

     Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9)

running 2 tests
test text_events_across_lines_keep_parent_and_literal_spans ... ok
test nested_and_mixed_wrappers_preserve_literal_titles ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

    Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-options/crates/markdown-view)
    Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.28s
   Compiling proc-macro2 v1.0.107
   Compiling unicode-ident v1.0.27
   Compiling quote v1.0.47
   Compiling memchr v2.8.3
   Compiling serde_core v1.0.229
   Compiling zmij v1.0.23
   Compiling aho-corasick v1.1.5
   Compiling regex-syntax v0.8.11
   Compiling syn v3.0.9
   Compiling serde_json v1.0.154
   Compiling pulldown-cmark v0.13.4 (/home/claude/Developer/Worktrees/calendar-805-options/vendor/pulldown-cmark)
   Compiling serde v1.0.229
   Compiling serde_derive v1.0.229
   Compiling regex-automata v0.4.18
   Compiling bitflags v2.13.2
   Compiling unicase v2.10.0
   Compiling itoa v1.0.18
   Compiling regex v1.13.1
   Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-options/crates/markdown-view)
    Finished `release` profile [optimized] target(s) in 42.49s

ADR 0003 evidence migration: `docs/plans/0067-slice2/options-build-evidence.txt` Source commit: [9064e29ac5ca9744cd369c1137a16bfd764e37d3](https://git.kayg.org/kayg/calternal/commit/9064e29ac5ca9744cd369c1137a16bfd764e37d3); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-options/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 0.63s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-7c85381e84b5f28c) running 3 tests test bounded::tests::deadline_is_exclusive ... ok test in_process::tests::hybrid_routes_boundaries_before_parsing ... ok test in_process::tests::unwind_and_limit_have_no_replacement_payload ... ok test result: ok. 3 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-950e03ded4713ee8) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-1b03582a2c4a86be) running 5 tests test existing_composer_configuration_matches_authored_titles ... ok test authored_link_contract ... ok test raw_configured_parser_label_evidence ... ok test original_line_contract_through_candidate ... ok test authored_source_rows ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/depth.rs (target/debug/deps/depth-e5a67ea44fd57263) running 5 tests test literal_markers_do_not_count_as_containers ... ok test hostile_depth_fails_without_partial_extraction ... ok test combined_depth_boundaries ... ok test multiline_list_depth_and_container_pops ... ok test seeded_mixed_containers_have_one_combined_budget ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/limits.rs (target/debug/deps/limits-7d451d6ff3157d32) running 4 tests test encoded_result_boundary_counts_all_json_bytes ... ok test every_count_boundary_is_exclusive ... ok test combined_seeded_fuzz_preserves_good_generation ... ok test default_source_and_node_boundaries ... ok test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.35s Running tests/tags.rs (target/debug/deps/tags-462dd1aff7d7abb9) running 2 tests test text_events_across_lines_keep_parent_and_literal_spans ... ok test nested_and_mixed_wrappers_preserve_literal_titles ... ok test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Checking calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-options/crates/markdown-view) Finished `dev` profile [unoptimized + debuginfo] target(s) in 0.28s Compiling proc-macro2 v1.0.107 Compiling unicode-ident v1.0.27 Compiling quote v1.0.47 Compiling memchr v2.8.3 Compiling serde_core v1.0.229 Compiling zmij v1.0.23 Compiling aho-corasick v1.1.5 Compiling regex-syntax v0.8.11 Compiling syn v3.0.9 Compiling serde_json v1.0.154 Compiling pulldown-cmark v0.13.4 (/home/claude/Developer/Worktrees/calendar-805-options/vendor/pulldown-cmark) Compiling serde v1.0.229 Compiling serde_derive v1.0.229 Compiling regex-automata v0.4.18 Compiling bitflags v2.13.2 Compiling unicase v2.10.0 Compiling itoa v1.0.18 Compiling regex v1.13.1 Compiling calternal-markdown-view v0.0.0 (/home/claude/Developer/Worktrees/calendar-805-options/crates/markdown-view) Finished `release` profile [optimized] target(s) in 42.49s ```

kayg commented

2026-11-02 12:42:02 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/options-memory-evidence.txt

Source commit: bda07f4763567f710ee37b30022edf610c6fece2; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

MEMORY-INPUT {"bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":55.157268,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
MEMORY-INPUT {"bytes": 2097152, "exit": 0, "stdout": "{\"elapsed_ms\":50.068961,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
MEMORY-INPUT {"bytes": 3145728, "exit": 0, "stdout": "{\"elapsed_ms\":50.122310000000006,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
MEMORY-INPUT {"bytes": 4194304, "exit": 0, "stdout": "{\"elapsed_ms\":50.242956,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
RSS-SAMPLE {"mode": "A", "exit": 0, "peak_simultaneous_rust_rss_kib": 8284, "individual_peak_rss_kib": {"25082": 8284}, "interval_ms": 5, "scope": "normal 10k fixture, not worst-case bound"}
RSS-SAMPLE {"mode": "C", "exit": 0, "peak_simultaneous_rust_rss_kib": 17180, "individual_peak_rss_kib": {"25086": 7968, "25089": 2544, "25090": 6980}, "interval_ms": 5, "scope": "normal 10k fixture, not worst-case bound"}

ADR 0003 evidence migration: `docs/plans/0067-slice2/options-memory-evidence.txt` Source commit: [bda07f4763567f710ee37b30022edf610c6fece2](https://git.kayg.org/kayg/calternal/commit/bda07f4763567f710ee37b30022edf610c6fece2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text MEMORY-INPUT {"bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":55.157268,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} MEMORY-INPUT {"bytes": 2097152, "exit": 0, "stdout": "{\"elapsed_ms\":50.068961,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} MEMORY-INPUT {"bytes": 3145728, "exit": 0, "stdout": "{\"elapsed_ms\":50.122310000000006,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} MEMORY-INPUT {"bytes": 4194304, "exit": 0, "stdout": "{\"elapsed_ms\":50.242956,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} RSS-SAMPLE {"mode": "A", "exit": 0, "peak_simultaneous_rust_rss_kib": 8284, "individual_peak_rss_kib": {"25082": 8284}, "interval_ms": 5, "scope": "normal 10k fixture, not worst-case bound"} RSS-SAMPLE {"mode": "C", "exit": 0, "peak_simultaneous_rust_rss_kib": 17180, "individual_peak_rss_kib": {"25086": 7968, "25089": 2544, "25090": 6980}, "interval_ms": 5, "scope": "normal 10k fixture, not worst-case bound"} ```

kayg commented

2026-11-02 12:42:02 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/options-perf-evidence.txt

FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033303000000000006,"p95_ms":0.04776,"p99_ms":0.056264999999999996},{"bytes":32768,"p50_ms":0.25536899999999996,"p95_ms":0.332352,"p99_ms":0.362991},{"bytes":262144,"p50_ms":2.540211,"p95_ms":2.859759,"p99_ms":3.024859}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.529539044}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033303000000000006, "p95_ms": 0.04776, "p99_ms": 0.056264999999999996}, {"bytes": 32768, "p50_ms": 0.25536899999999996, "p95_ms": 0.332352, "p99_ms": 0.362991}, {"bytes": 262144, "p50_ms": 2.540211, "p95_ms": 2.859759, "p99_ms": 3.024859}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.529539044, "cpu_s": 0.8288589999999999, "maxrss_kib": 20120}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.092323,"p95_ms":0.114315,"p99_ms":0.127719},{"bytes":32768,"p50_ms":0.63455,"p95_ms":0.729377,"p99_ms":0.8227920000000001},{"bytes":262144,"p50_ms":5.312736999999999,"p95_ms":5.692257,"p99_ms":5.889718}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":1.222194403}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.092323, "p95_ms": 0.114315, "p99_ms": 0.127719}, {"bytes": 32768, "p50_ms": 0.63455, "p95_ms": 0.729377, "p99_ms": 0.8227920000000001}, {"bytes": 262144, "p50_ms": 5.312736999999999, "p95_ms": 5.692257, "p99_ms": 5.889718}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 1.222194403, "cpu_s": 1.989572, "maxrss_kib": 20120}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.10149000000000001,"p95_ms":0.13316,"p99_ms":0.149581},{"bytes":32768,"p50_ms":0.706043,"p95_ms":0.820247,"p99_ms":0.898845},{"bytes":262144,"p50_ms":6.484652,"p95_ms":7.512689,"p99_ms":10.594233999999998}],"bytes":92569600,"edges":5000,"isolated":100,"mode":"C","notes":10000,"rows":20000,"unresolved":0,"wall_s":1.47664325}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.10149000000000001, "p95_ms": 0.13316, "p99_ms": 0.149581}, {"bytes": 32768, "p50_ms": 0.706043, "p95_ms": 0.820247, "p99_ms": 0.898845}, {"bytes": 262144, "p50_ms": 6.484652, "p95_ms": 7.512689, "p99_ms": 10.594233999999998}], "bytes": 92569600, "edges": 5000, "isolated": 100, "mode": "C", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 1.47664325, "cpu_s": 2.367851, "maxrss_kib": 20120}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033803999999999994,"p95_ms":0.049352,"p99_ms":0.058911},{"bytes":32768,"p50_ms":0.260998,"p95_ms":0.35481500000000005,"p99_ms":0.385442},{"bytes":262144,"p50_ms":2.594563,"p95_ms":3.38886,"p99_ms":4.069697}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.541010731}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033803999999999994, "p95_ms": 0.049352, "p99_ms": 0.058911}, {"bytes": 32768, "p50_ms": 0.260998, "p95_ms": 0.35481500000000005, "p99_ms": 0.385442}, {"bytes": 262144, "p50_ms": 2.594563, "p95_ms": 3.38886, "p99_ms": 4.069697}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.541010731, "cpu_s": 8.658325999999999, "maxrss_kib": 20120}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.093365,"p95_ms":0.115857,"p99_ms":0.130736},{"bytes":32768,"p50_ms":0.6389670000000001,"p95_ms":0.754764,"p99_ms":0.82712},{"bytes":262144,"p50_ms":5.370244,"p95_ms":6.1073960000000005,"p99_ms":6.436572}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":12.437637394}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.093365, "p95_ms": 0.115857, "p99_ms": 0.130736}, {"bytes": 32768, "p50_ms": 0.6389670000000001, "p95_ms": 0.754764, "p99_ms": 0.82712}, {"bytes": 262144, "p50_ms": 5.370244, "p95_ms": 6.1073960000000005, "p99_ms": 6.436572}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 12.437637394, "cpu_s": 20.182065, "maxrss_kib": 20120}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.101089,"p95_ms":0.129894,"p99_ms":0.147957},{"bytes":32768,"p50_ms":0.7040599999999999,"p95_ms":0.841597,"p99_ms":0.9390499999999999},{"bytes":262144,"p50_ms":6.3140339999999995,"p95_ms":8.12161,"p99_ms":10.66668}],"bytes":925696000,"edges":50000,"isolated":1000,"mode":"C","notes":100000,"rows":200000,"unresolved":0,"wall_s":14.3346124}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.101089, "p95_ms": 0.129894, "p99_ms": 0.147957}, {"bytes": 32768, "p50_ms": 0.7040599999999999, "p95_ms": 0.841597, "p99_ms": 0.9390499999999999}, {"bytes": 262144, "p50_ms": 6.3140339999999995, "p95_ms": 8.12161, "p99_ms": 10.66668}], "bytes": 925696000, "edges": 50000, "isolated": 1000, "mode": "C", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 14.3346124, "cpu_s": 22.965857, "maxrss_kib": 20120}
BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.06650500000000001, "growth": 1.3475644350786191, "pass": false}, {"bytes": 32768, "added_p95_ms": 0.39994899999999994, "growth": 1.127204317743049, "pass": false}, {"bytes": 262144, "added_p95_ms": 2.7185360000000003, "growth": 0.8021977892270558, "pass": false}], "vs_historical_12.614s": -0.013981497225305284, "vs_paired_direct": 1.2446513818166434, "historical_wall_pass": true, "paired_wall_pass": false}
BUDGET {"option": "C", "buckets": [{"bytes": 4096, "added_p95_ms": 0.080542, "growth": 1.6319905981520506, "pass": false}, {"bytes": 32768, "added_p95_ms": 0.486782, "growth": 1.371931851810098, "pass": false}, {"bytes": 262144, "added_p95_ms": 4.73275, "growth": 1.3965610854387611, "pass": false}], "vs_historical_12.614s": 0.1364049785952115, "vs_paired_direct": 1.587003183336733, "historical_wall_pass": false, "paired_wall_pass": false}
CONTAINMENT {"probe": "ordinary", "bytes": 24, "exit": 0, "stdout": "{\"elapsed_ms\":1.65439,\"outcome\":{\"rows\":2,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "depth255", "bytes": 518, "exit": 0, "stdout": "{\"elapsed_ms\":1.449316,\"outcome\":{\"rows\":1,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "depth256", "bytes": 520, "exit": 0, "stdout": "{\"elapsed_ms\":0.024195,\"outcome\":{\"reason\":\"depth-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "long-paragraph", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":6.553742000000001,\"outcome\":{\"rows\":0,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "short-line-storm", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":50.503195,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "inline-storm", "bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":51.095185,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "source-limit", "bytes": 8388608, "exit": 0, "stdout": "{\"elapsed_ms\":0.70436,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}

ADR 0003 evidence migration: `docs/plans/0067-slice2/options-perf-evidence.txt` Source commit: [bda07f4763567f710ee37b30022edf610c6fece2](https://git.kayg.org/kayg/calternal/commit/bda07f4763567f710ee37b30022edf610c6fece2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033303000000000006,"p95_ms":0.04776,"p99_ms":0.056264999999999996},{"bytes":32768,"p50_ms":0.25536899999999996,"p95_ms":0.332352,"p99_ms":0.362991},{"bytes":262144,"p50_ms":2.540211,"p95_ms":2.859759,"p99_ms":3.024859}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.529539044} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033303000000000006, "p95_ms": 0.04776, "p99_ms": 0.056264999999999996}, {"bytes": 32768, "p50_ms": 0.25536899999999996, "p95_ms": 0.332352, "p99_ms": 0.362991}, {"bytes": 262144, "p50_ms": 2.540211, "p95_ms": 2.859759, "p99_ms": 3.024859}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.529539044, "cpu_s": 0.8288589999999999, "maxrss_kib": 20120} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.092323,"p95_ms":0.114315,"p99_ms":0.127719},{"bytes":32768,"p50_ms":0.63455,"p95_ms":0.729377,"p99_ms":0.8227920000000001},{"bytes":262144,"p50_ms":5.312736999999999,"p95_ms":5.692257,"p99_ms":5.889718}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":1.222194403} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.092323, "p95_ms": 0.114315, "p99_ms": 0.127719}, {"bytes": 32768, "p50_ms": 0.63455, "p95_ms": 0.729377, "p99_ms": 0.8227920000000001}, {"bytes": 262144, "p50_ms": 5.312736999999999, "p95_ms": 5.692257, "p99_ms": 5.889718}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 1.222194403, "cpu_s": 1.989572, "maxrss_kib": 20120} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.10149000000000001,"p95_ms":0.13316,"p99_ms":0.149581},{"bytes":32768,"p50_ms":0.706043,"p95_ms":0.820247,"p99_ms":0.898845},{"bytes":262144,"p50_ms":6.484652,"p95_ms":7.512689,"p99_ms":10.594233999999998}],"bytes":92569600,"edges":5000,"isolated":100,"mode":"C","notes":10000,"rows":20000,"unresolved":0,"wall_s":1.47664325} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.10149000000000001, "p95_ms": 0.13316, "p99_ms": 0.149581}, {"bytes": 32768, "p50_ms": 0.706043, "p95_ms": 0.820247, "p99_ms": 0.898845}, {"bytes": 262144, "p50_ms": 6.484652, "p95_ms": 7.512689, "p99_ms": 10.594233999999998}], "bytes": 92569600, "edges": 5000, "isolated": 100, "mode": "C", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 1.47664325, "cpu_s": 2.367851, "maxrss_kib": 20120} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033803999999999994,"p95_ms":0.049352,"p99_ms":0.058911},{"bytes":32768,"p50_ms":0.260998,"p95_ms":0.35481500000000005,"p99_ms":0.385442},{"bytes":262144,"p50_ms":2.594563,"p95_ms":3.38886,"p99_ms":4.069697}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.541010731} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033803999999999994, "p95_ms": 0.049352, "p99_ms": 0.058911}, {"bytes": 32768, "p50_ms": 0.260998, "p95_ms": 0.35481500000000005, "p99_ms": 0.385442}, {"bytes": 262144, "p50_ms": 2.594563, "p95_ms": 3.38886, "p99_ms": 4.069697}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.541010731, "cpu_s": 8.658325999999999, "maxrss_kib": 20120} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.093365,"p95_ms":0.115857,"p99_ms":0.130736},{"bytes":32768,"p50_ms":0.6389670000000001,"p95_ms":0.754764,"p99_ms":0.82712},{"bytes":262144,"p50_ms":5.370244,"p95_ms":6.1073960000000005,"p99_ms":6.436572}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":12.437637394} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.093365, "p95_ms": 0.115857, "p99_ms": 0.130736}, {"bytes": 32768, "p50_ms": 0.6389670000000001, "p95_ms": 0.754764, "p99_ms": 0.82712}, {"bytes": 262144, "p50_ms": 5.370244, "p95_ms": 6.1073960000000005, "p99_ms": 6.436572}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 12.437637394, "cpu_s": 20.182065, "maxrss_kib": 20120} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.101089,"p95_ms":0.129894,"p99_ms":0.147957},{"bytes":32768,"p50_ms":0.7040599999999999,"p95_ms":0.841597,"p99_ms":0.9390499999999999},{"bytes":262144,"p50_ms":6.3140339999999995,"p95_ms":8.12161,"p99_ms":10.66668}],"bytes":925696000,"edges":50000,"isolated":1000,"mode":"C","notes":100000,"rows":200000,"unresolved":0,"wall_s":14.3346124} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.101089, "p95_ms": 0.129894, "p99_ms": 0.147957}, {"bytes": 32768, "p50_ms": 0.7040599999999999, "p95_ms": 0.841597, "p99_ms": 0.9390499999999999}, {"bytes": 262144, "p50_ms": 6.3140339999999995, "p95_ms": 8.12161, "p99_ms": 10.66668}], "bytes": 925696000, "edges": 50000, "isolated": 1000, "mode": "C", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 14.3346124, "cpu_s": 22.965857, "maxrss_kib": 20120} BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.06650500000000001, "growth": 1.3475644350786191, "pass": false}, {"bytes": 32768, "added_p95_ms": 0.39994899999999994, "growth": 1.127204317743049, "pass": false}, {"bytes": 262144, "added_p95_ms": 2.7185360000000003, "growth": 0.8021977892270558, "pass": false}], "vs_historical_12.614s": -0.013981497225305284, "vs_paired_direct": 1.2446513818166434, "historical_wall_pass": true, "paired_wall_pass": false} BUDGET {"option": "C", "buckets": [{"bytes": 4096, "added_p95_ms": 0.080542, "growth": 1.6319905981520506, "pass": false}, {"bytes": 32768, "added_p95_ms": 0.486782, "growth": 1.371931851810098, "pass": false}, {"bytes": 262144, "added_p95_ms": 4.73275, "growth": 1.3965610854387611, "pass": false}], "vs_historical_12.614s": 0.1364049785952115, "vs_paired_direct": 1.587003183336733, "historical_wall_pass": false, "paired_wall_pass": false} CONTAINMENT {"probe": "ordinary", "bytes": 24, "exit": 0, "stdout": "{\"elapsed_ms\":1.65439,\"outcome\":{\"rows\":2,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "depth255", "bytes": 518, "exit": 0, "stdout": "{\"elapsed_ms\":1.449316,\"outcome\":{\"rows\":1,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "depth256", "bytes": 520, "exit": 0, "stdout": "{\"elapsed_ms\":0.024195,\"outcome\":{\"reason\":\"depth-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "long-paragraph", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":6.553742000000001,\"outcome\":{\"rows\":0,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "short-line-storm", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":50.503195,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "inline-storm", "bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":51.095185,\"outcome\":{\"reason\":\"timeout\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "source-limit", "bytes": 8388608, "exit": 0, "stdout": "{\"elapsed_ms\":0.70436,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} ```

kayg commented

2026-11-02 12:42:03 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/package-arm-evidence.txt

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
   Compiling proc-macro2 v1.0.107
   Compiling memchr v2.8.3
   Compiling quote v1.0.47
   Compiling unicode-ident v1.0.27
   Compiling serde_core v1.0.229
   Compiling zmij v1.0.23
   Compiling aho-corasick v1.1.5
   Compiling syn v3.0.9
   Compiling serde v1.0.229
   Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-805-all-gates/vendor/pulldown-cmark)
   Compiling regex-syntax v0.8.11
   Compiling serde_json v1.0.154
   Compiling bitflags v2.13.2
   Compiling unicase v2.10.0
   Compiling itoa v1.0.18
   Compiling regex-automata v0.4.18
   Compiling regex v1.13.1
   Compiling serde_derive v1.0.229
   Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-805-all-gates/crates/markdown-view)
    Finished `test` profile [unoptimized + debuginfo] target(s) in 9.51s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-ba4ab689e69bdd9f)

running 1 test
test bounded::tests::deadline_is_exclusive ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-c51dc955e741279c)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-9fa429d3a450f935)

running 5 tests
test existing_composer_configuration_matches_authored_titles ... ok
test authored_link_contract ... ok
test raw_configured_parser_label_evidence ... ok
test authored_source_rows ... ok
test original_line_contract_through_candidate ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/depth.rs (target/debug/deps/depth-b9eedd1ae57bf3e0)

running 5 tests
test hostile_depth_fails_without_partial_extraction ... ok
test literal_markers_do_not_count_as_containers ... ok
test combined_depth_boundaries ... ok
test multiline_list_depth_and_container_pops ... ok
test seeded_mixed_containers_have_one_combined_budget ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/limits.rs (target/debug/deps/limits-ca371733c847007c)

running 4 tests
test encoded_result_boundary_counts_all_json_bytes ... ok
test every_count_boundary_is_exclusive ... ok
test combined_seeded_fuzz_preserves_good_generation ... ok
test default_source_and_node_boundaries ... ok

test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.15s

     Running tests/tags.rs (target/debug/deps/tags-be184587f6a0da07)

running 2 tests
test text_events_across_lines_keep_parent_and_literal_spans ... ok
test nested_and_mixed_wrappers_preserve_literal_titles ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

   Compiling proc-macro2 v1.0.107
   Compiling quote v1.0.47
   Compiling memchr v2.8.3
   Compiling unicode-ident v1.0.27
   Compiling serde_core v1.0.229
   Compiling zmij v1.0.23
   Compiling aho-corasick v1.1.5
   Compiling syn v3.0.9
   Compiling serde_json v1.0.154
   Compiling regex-syntax v0.8.11
   Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-805-all-gates/vendor/pulldown-cmark)
   Compiling serde v1.0.229
   Compiling serde_derive v1.0.229
   Compiling itoa v1.0.18
   Compiling regex-automata v0.4.18
   Compiling unicase v2.10.0
   Compiling bitflags v2.13.2
   Compiling regex v1.13.1
   Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-805-all-gates/crates/markdown-view)
    Finished `release` profile [optimized] target(s) in 8.86s
1d880a9ec906c6171fa4b105b1fac5fc1adcb00379fef4ef84b31adb2369ef1a  target/release/markdown-worker
evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
{"os": "Linux", "arch": "aarch64", "ordinary": "pass", "oom": "pass", "stack": "pass"}

ADR 0003 evidence migration: `docs/plans/0067-slice2/package-arm-evidence.txt` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead Compiling proc-macro2 v1.0.107 Compiling memchr v2.8.3 Compiling quote v1.0.47 Compiling unicode-ident v1.0.27 Compiling serde_core v1.0.229 Compiling zmij v1.0.23 Compiling aho-corasick v1.1.5 Compiling syn v3.0.9 Compiling serde v1.0.229 Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-805-all-gates/vendor/pulldown-cmark) Compiling regex-syntax v0.8.11 Compiling serde_json v1.0.154 Compiling bitflags v2.13.2 Compiling unicase v2.10.0 Compiling itoa v1.0.18 Compiling regex-automata v0.4.18 Compiling regex v1.13.1 Compiling serde_derive v1.0.229 Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-805-all-gates/crates/markdown-view) Finished `test` profile [unoptimized + debuginfo] target(s) in 9.51s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-ba4ab689e69bdd9f) running 1 test test bounded::tests::deadline_is_exclusive ... ok test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running unittests src/bin/markdown-worker.rs (target/debug/deps/markdown_worker-c51dc955e741279c) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-9fa429d3a450f935) running 5 tests test existing_composer_configuration_matches_authored_titles ... ok test authored_link_contract ... ok test raw_configured_parser_label_evidence ... ok test authored_source_rows ... ok test original_line_contract_through_candidate ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/depth.rs (target/debug/deps/depth-b9eedd1ae57bf3e0) running 5 tests test hostile_depth_fails_without_partial_extraction ... ok test literal_markers_do_not_count_as_containers ... ok test combined_depth_boundaries ... ok test multiline_list_depth_and_container_pops ... ok test seeded_mixed_containers_have_one_combined_budget ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Running tests/limits.rs (target/debug/deps/limits-ca371733c847007c) running 4 tests test encoded_result_boundary_counts_all_json_bytes ... ok test every_count_boundary_is_exclusive ... ok test combined_seeded_fuzz_preserves_good_generation ... ok test default_source_and_node_boundaries ... ok test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.15s Running tests/tags.rs (target/debug/deps/tags-be184587f6a0da07) running 2 tests test text_events_across_lines_keep_parent_and_literal_spans ... ok test nested_and_mixed_wrappers_preserve_literal_titles ... ok test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Compiling proc-macro2 v1.0.107 Compiling quote v1.0.47 Compiling memchr v2.8.3 Compiling unicode-ident v1.0.27 Compiling serde_core v1.0.229 Compiling zmij v1.0.23 Compiling aho-corasick v1.1.5 Compiling syn v3.0.9 Compiling serde_json v1.0.154 Compiling regex-syntax v0.8.11 Compiling pulldown-cmark v0.13.4 (/home/cua/Developer/Worktrees/calendar-805-all-gates/vendor/pulldown-cmark) Compiling serde v1.0.229 Compiling serde_derive v1.0.229 Compiling itoa v1.0.18 Compiling regex-automata v0.4.18 Compiling unicase v2.10.0 Compiling bitflags v2.13.2 Compiling regex v1.13.1 Compiling calternal-markdown-view v0.0.0 (/home/cua/Developer/Worktrees/calendar-805-all-gates/crates/markdown-view) Finished `release` profile [optimized] target(s) in 8.86s 1d880a9ec906c6171fa4b105b1fac5fc1adcb00379fef4ef84b31adb2369ef1a target/release/markdown-worker evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead {"os": "Linux", "arch": "aarch64", "ordinary": "pass", "oom": "pass", "stack": "pass"} ```

kayg commented

2026-11-02 12:42:03 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/package-mac-evidence.txt

mac packaging cancelled at measured budget failure; never acquired shared mac lock

ADR 0003 evidence migration: `docs/plans/0067-slice2/package-mac-evidence.txt` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text mac packaging cancelled at measured budget failure; never acquired shared mac lock ```

kayg commented

2026-11-02 12:42:03 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/package-x86-evidence.txt

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
{"os": "Linux", "arch": "x86_64", "ordinary": "pass", "oom": "pass", "stack": "pass"}

ADR 0003 evidence migration: `docs/plans/0067-slice2/package-x86-evidence.txt` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead {"os": "Linux", "arch": "x86_64", "ordinary": "pass", "oom": "pass", "stack": "pass"} ```

kayg commented

2026-11-02 12:42:04 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/resume-evidence.txt

Source commit: 5e07725121a50980784aaac27ef92b807873f633; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

Fresh VM evidence at 2da88cbd1bfe8ba7a81ffc6f014b5724725e7c23
Host: claude@calternal-dev-01-compute02; clean worktree calendar-v2-s2-805-verify
Commands and stop decision: README.md
Build chatter and repetitive successful vector names omitted; assertions unchanged.

     Running tests/contract.rs (target/debug/deps/contract-366ef1b510a86bf3)

running 5 tests
13/13 authored link vectors passed
test authored_link_contract ... ok
13/13 untouched Composer configuration titles match authored fixtures
[H~2~O](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("H~2~O")), End(Link), End(Paragraph)]
[x^2^](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("x^2")), Text(Borrowed("^")), End(Link), End(Paragraph)]
[~~old~~](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Start(Strikethrough), Text(Borrowed("old")), End(Strikethrough), End(Link), End(Paragraph)]
test existing_composer_configuration_matches_authored_titles ... ok
test raw_configured_parser_label_evidence ... ok

thread 'authored_source_rows' (3083163) panicked at crates/markdown-view/tests/contract.rs:113:5:
"formatted-tag-source-boundary": got {"classification":["e"],"relationships":[],"rows":[{"kind":"event","line":0,"tags":[],"time":{"source":"09:00","span":[2,7]},"title":[{"source":"Meet **#area/work**","span":[8,27]}]}]}, expected {"classification":["e"],"relationships":[],"rows":[{"kind":"event","line":0,"tags":[{"source":"#area/work","span":[15,25]}],"time":{"source":"09:00","span":[2,7]},"title":[{"source":"Meet","span":[8,12]}]}]}
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
test authored_source_rows ... FAILED
test original_line_contract_through_candidate ... ok

failures:

failures:
    authored_source_rows

test result: FAILED. 4 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

error: test failed, to rerun pass `-p calternal-markdown-view --test contract`

     Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef)

running 1 test
70/70 pulldown line vectors passed
test tests::shared_line_vectors ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

   Doc-tests calendar_0067_parity_spike

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running unittests src/lib.rs (target/debug/deps/calternal_tasks-1cfffd559dae52d1)

running 12 tests
composer "ordinary-source": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "composer-person-link": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "unicode-byte-span": [240, 159, 152, 128, 32, 99, 97, 102, 195, 169]
test create_tests::concurrent_creates_name_clash_and_new_daily_undo ... ok
composer "composer-formatting-outside-link": [42, 42, 77, 101, 101, 116, 42, 42, 32, 80, 114, 105, 121, 97, 32, 116, 111, 100, 97, 121]
composer "composer-emphasis-in-label": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "composer-code-in-label": [77, 101, 101, 116, 32, 97, 32, 98]
test create_tests::create_ack_can_be_undone_immediately_and_reuses_daily_placeholder ... ok
composer "code-does-not-create-link": [96, 91, 80, 114, 105, 121, 97, 93, 40, 60, 112, 46, 109, 100, 62, 41, 96]
composer "two-links": [65, 32, 97, 110, 100, 32, 66]
test create_tests::create_isolation_and_idempotent_retry ... ok
composer "emoji-source-label": [77, 101, 101, 116, 32, 58, 117, 110, 107, 110, 111, 119, 110, 95, 78, 65, 77, 69, 58]
composer "composer-subscript-compatibility": [77, 101, 101, 116, 32, 72, 126, 50, 126, 79]
composer "composer-superscript-compatibility": [77, 101, 101, 116, 32, 120, 94, 50, 94]
composer "composer-strikethrough-compatibility": [77, 101, 101, 116, 32, 126, 126, 111, 108, 100, 126, 126]
test create_tests::daily_uses_device_offset_and_invalid_slots_write_nothing ... ok
test create_tests::composer_authored_title_corpus ... ok
test create_tests::only_links_are_reduced_in_the_note_title ... ok
test create_tests::create_and_undo_restore_daily_bytes_and_remove_task ... ok
test create_tests::description_heading_keeps_the_given_task_title ... ok
test create_tests::opening_today_then_creating_tasks_scopes_keys_and_retries ... ok
test create_tests::undo_new_daily_capture_keeps_later_prose_and_removes_its_lines ... ok
test create_tests::repeated_create_undo_recreates_daily_with_retained_history ... ok
test create_tests::create_and_undo_one_mb_daily ... ok

test result: ok. 12 passed; 0 failed; 0 ignored; 0 measured; 30 filtered out; finished in 1.29s

 RUN  v5.0.4 /home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/packages/editor

 ✓ test/markdown-contract.test.ts (13 tests) 41ms
 ✓ test/calendar-0067.spike.test.ts (70 tests) 74ms
 ❯ test/markdown-rows.test.ts (96 tests | 2 failed) 143ms
   ✓ authored source rows: literal-title-0 21ms
   ✓ authored source rows: literal-title-1 2ms
   ✓ authored source rows: literal-title-2 2ms
   ✓ authored source rows: literal-title-3 1ms
   ✓ authored source rows: literal-title-4 3ms
   ✓ authored source rows: literal-title-5 2ms
   ✓ authored source rows: literal-title-6 2ms
   ✓ authored source rows: literal-title-7 3ms
   ✓ authored source rows: literal-title-8 4ms
   ✓ authored source rows: utf8-task-crlf 3ms
   ✓ authored source rows: html-sub-title 3ms
   ✓ authored source rows: html-sup-title 1ms
   ✓ authored source rows: multiline-code 3ms
   ✓ authored source rows: fence 1ms
   ✓ authored source rows: indented 1ms
   ✓ authored source rows: html-comment 1ms
   ✓ authored source rows: markdown-comment 1ms
   ✓ authored source rows: frontmatter 4ms
   ✓ authored source rows: direct-children 2ms
   ✓ authored source rows: commented-relationship 2ms
   ✓ authored source rows: every-ancestor-event 2ms
   ✓ authored source rows: quoted-checkbox-child 2ms
   ✓ authored source rows: sentence-not-relationship 2ms
   ✓ authored source rows: list-boundary-ends-code-span 2ms
   × authored source rows: formatted-tag-source-boundary 21ms
   × authored source rows: literal-interior-bom-title 4ms

⎯⎯⎯⎯⎯⎯⎯ Failed Tests 2 ⎯⎯⎯⎯⎯⎯⎯

 FAIL  test/markdown-rows.test.ts > authored source rows: formatted-tag-source-boundary
AssertionError: expected { classification: [ 'e' ], …(2) } to deeply equal { classification: [ 'e' ], …(2) }

- Expected
+ Received

@@ -5,32 +5,24 @@
    "relationships": [],
    "rows": [
      {
        "kind": "event",
        "line": 0,
-       "tags": [
-         {
-           "source": "#area/work",
-           "span": [
-             15,
-             25,
-           ],
-         },
-       ],
+       "tags": [],
        "time": {
          "source": "09:00",
          "span": [
            2,
            7,
          ],
        },
        "title": [
          {
-           "source": "Meet",
+           "source": "Meet **#area/work**",
            "span": [
              8,
-             12,
+             27,
            ],
          },
        ],
      },
    ],

 ❯ test/markdown-rows.test.ts:26:15
     24|     return value
     25|   }))
     26|   expect(got).toEqual(v.expected)
       |               ^
     27| })
     28| for (const v of original.vectors) it(`candidate original line vector: …

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[1/2]⎯

 FAIL  test/markdown-rows.test.ts > authored source rows: literal-interior-bom-title
AssertionError: expected { classification: [ 'e' ], …(2) } to deeply equal { classification: [ 'e' ], …(2) }

- Expected
+ Received

@@ -15,14 +15,14 @@
            7,
          ],
        },
        "title": [
          {
-           "source": "Meet",
+           "source": "Meet",
            "span": [
              8,
-             15,
+             12,
            ],
          },
        ],
      },
    ],

 ❯ test/markdown-rows.test.ts:26:15
     24|     return value
     25|   }))
     26|   expect(got).toEqual(v.expected)
       |               ^
     27| })
     28| for (const v of original.vectors) it(`candidate original line vector: …

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[2/2]⎯


 Test Files  1 failed | 2 passed (3)
      Tests  2 failed | 177 passed (179)
   Start at  19:53:43
   Duration  1.34s (environment 50%, import 17%, transform 16%, tests 8%, setup 7%, worker 2%)

$ tsc --noEmit
GATE_STATUS native=101 seed=0 composer=0 install=0 editor=1 types=0

Composer byte comparison:
recovered before == recovered after == fresh 2da88cbd: 12/12 Composer byte arrays identical
composer "code-does-not-create-link": [96, 91, 80, 114, 105, 121, 97, 93, 40, 60, 112, 46, 109, 100, 62, 41, 96]
composer "composer-code-in-label": [77, 101, 101, 116, 32, 97, 32, 98]
composer "composer-emphasis-in-label": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "composer-formatting-outside-link": [42, 42, 77, 101, 101, 116, 42, 42, 32, 80, 114, 105, 121, 97, 32, 116, 111, 100, 97, 121]
composer "composer-person-link": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "composer-strikethrough-compatibility": [77, 101, 101, 116, 32, 126, 126, 111, 108, 100, 126, 126]
composer "composer-subscript-compatibility": [77, 101, 101, 116, 32, 72, 126, 50, 126, 79]
composer "composer-superscript-compatibility": [77, 101, 101, 116, 32, 120, 94, 50, 94]
composer "emoji-source-label": [77, 101, 101, 116, 32, 58, 117, 110, 107, 110, 111, 119, 110, 95, 78, 65, 77, 69, 58]
composer "ordinary-source": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "two-links": [65, 32, 97, 110, 100, 32, 66]
composer "unicode-byte-span": [240, 159, 152, 128, 32, 99, 97, 102, 195, 169]

Corpus SHA256:
b9d3635b6eb121610e5aefe54492144f526b38c1bf50902af111c96fb096632e  contracts/markdown-extraction/composer.json
8346eaf880bbb8940bf642e27f473fe639caf1017e8b5254192fd44b1e9ca838  contracts/markdown-extraction/links.json
94afdf93c5f61b6821334324794512c5a62d5f1b52799d56c018200b0ae69c3a  contracts/markdown-extraction/rows.json

ADR 0003 evidence migration: `docs/plans/0067-slice2/resume-evidence.txt` Source commit: [5e07725121a50980784aaac27ef92b807873f633](https://git.kayg.org/kayg/calternal/commit/5e07725121a50980784aaac27ef92b807873f633); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text Fresh VM evidence at 2da88cbd1bfe8ba7a81ffc6f014b5724725e7c23 Host: claude@calternal-dev-01-compute02; clean worktree calendar-v2-s2-805-verify Commands and stop decision: README.md Build chatter and repetitive successful vector names omitted; assertions unchanged. Running tests/contract.rs (target/debug/deps/contract-366ef1b510a86bf3) running 5 tests 13/13 authored link vectors passed test authored_link_contract ... ok 13/13 untouched Composer configuration titles match authored fixtures [H~2~O](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("H~2~O")), End(Link), End(Paragraph)] [x^2^](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("x^2")), Text(Borrowed("^")), End(Link), End(Paragraph)] [~~old~~](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Start(Strikethrough), Text(Borrowed("old")), End(Strikethrough), End(Link), End(Paragraph)] test existing_composer_configuration_matches_authored_titles ... ok test raw_configured_parser_label_evidence ... ok thread 'authored_source_rows' (3083163) panicked at crates/markdown-view/tests/contract.rs:113:5: "formatted-tag-source-boundary": got {"classification":["e"],"relationships":[],"rows":[{"kind":"event","line":0,"tags":[],"time":{"source":"09:00","span":[2,7]},"title":[{"source":"Meet **#area/work**","span":[8,27]}]}]}, expected {"classification":["e"],"relationships":[],"rows":[{"kind":"event","line":0,"tags":[{"source":"#area/work","span":[15,25]}],"time":{"source":"09:00","span":[2,7]},"title":[{"source":"Meet","span":[8,12]}]}]} note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace test authored_source_rows ... FAILED test original_line_contract_through_candidate ... ok failures: failures: authored_source_rows test result: FAILED. 4 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s error: test failed, to rerun pass `-p calternal-markdown-view --test contract` Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef) running 1 test 70/70 pulldown line vectors passed test tests::shared_line_vectors ... ok test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Doc-tests calendar_0067_parity_spike running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running unittests src/lib.rs (target/debug/deps/calternal_tasks-1cfffd559dae52d1) running 12 tests composer "ordinary-source": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "composer-person-link": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "unicode-byte-span": [240, 159, 152, 128, 32, 99, 97, 102, 195, 169] test create_tests::concurrent_creates_name_clash_and_new_daily_undo ... ok composer "composer-formatting-outside-link": [42, 42, 77, 101, 101, 116, 42, 42, 32, 80, 114, 105, 121, 97, 32, 116, 111, 100, 97, 121] composer "composer-emphasis-in-label": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "composer-code-in-label": [77, 101, 101, 116, 32, 97, 32, 98] test create_tests::create_ack_can_be_undone_immediately_and_reuses_daily_placeholder ... ok composer "code-does-not-create-link": [96, 91, 80, 114, 105, 121, 97, 93, 40, 60, 112, 46, 109, 100, 62, 41, 96] composer "two-links": [65, 32, 97, 110, 100, 32, 66] test create_tests::create_isolation_and_idempotent_retry ... ok composer "emoji-source-label": [77, 101, 101, 116, 32, 58, 117, 110, 107, 110, 111, 119, 110, 95, 78, 65, 77, 69, 58] composer "composer-subscript-compatibility": [77, 101, 101, 116, 32, 72, 126, 50, 126, 79] composer "composer-superscript-compatibility": [77, 101, 101, 116, 32, 120, 94, 50, 94] composer "composer-strikethrough-compatibility": [77, 101, 101, 116, 32, 126, 126, 111, 108, 100, 126, 126] test create_tests::daily_uses_device_offset_and_invalid_slots_write_nothing ... ok test create_tests::composer_authored_title_corpus ... ok test create_tests::only_links_are_reduced_in_the_note_title ... ok test create_tests::create_and_undo_restore_daily_bytes_and_remove_task ... ok test create_tests::description_heading_keeps_the_given_task_title ... ok test create_tests::opening_today_then_creating_tasks_scopes_keys_and_retries ... ok test create_tests::undo_new_daily_capture_keeps_later_prose_and_removes_its_lines ... ok test create_tests::repeated_create_undo_recreates_daily_with_retained_history ... ok test create_tests::create_and_undo_one_mb_daily ... ok test result: ok. 12 passed; 0 failed; 0 ignored; 0 measured; 30 filtered out; finished in 1.29s RUN v5.0.4 /home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/packages/editor ✓ test/markdown-contract.test.ts (13 tests) 41ms ✓ test/calendar-0067.spike.test.ts (70 tests) 74ms ❯ test/markdown-rows.test.ts (96 tests | 2 failed) 143ms ✓ authored source rows: literal-title-0 21ms ✓ authored source rows: literal-title-1 2ms ✓ authored source rows: literal-title-2 2ms ✓ authored source rows: literal-title-3 1ms ✓ authored source rows: literal-title-4 3ms ✓ authored source rows: literal-title-5 2ms ✓ authored source rows: literal-title-6 2ms ✓ authored source rows: literal-title-7 3ms ✓ authored source rows: literal-title-8 4ms ✓ authored source rows: utf8-task-crlf 3ms ✓ authored source rows: html-sub-title 3ms ✓ authored source rows: html-sup-title 1ms ✓ authored source rows: multiline-code 3ms ✓ authored source rows: fence 1ms ✓ authored source rows: indented 1ms ✓ authored source rows: html-comment 1ms ✓ authored source rows: markdown-comment 1ms ✓ authored source rows: frontmatter 4ms ✓ authored source rows: direct-children 2ms ✓ authored source rows: commented-relationship 2ms ✓ authored source rows: every-ancestor-event 2ms ✓ authored source rows: quoted-checkbox-child 2ms ✓ authored source rows: sentence-not-relationship 2ms ✓ authored source rows: list-boundary-ends-code-span 2ms × authored source rows: formatted-tag-source-boundary 21ms × authored source rows: literal-interior-bom-title 4ms ⎯⎯⎯⎯⎯⎯⎯ Failed Tests 2 ⎯⎯⎯⎯⎯⎯⎯ FAIL test/markdown-rows.test.ts > authored source rows: formatted-tag-source-boundary AssertionError: expected { classification: [ 'e' ], …(2) } to deeply equal { classification: [ 'e' ], …(2) } - Expected + Received @@ -5,32 +5,24 @@ "relationships": [], "rows": [ { "kind": "event", "line": 0, - "tags": [ - { - "source": "#area/work", - "span": [ - 15, - 25, - ], - }, - ], + "tags": [], "time": { "source": "09:00", "span": [ 2, 7, ], }, "title": [ { - "source": "Meet", + "source": "Meet **#area/work**", "span": [ 8, - 12, + 27, ], }, ], }, ], ❯ test/markdown-rows.test.ts:26:15 24| return value 25| })) 26| expect(got).toEqual(v.expected) | ^ 27| }) 28| for (const v of original.vectors) it(`candidate original line vector: … ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[1/2]⎯ FAIL test/markdown-rows.test.ts > authored source rows: literal-interior-bom-title AssertionError: expected { classification: [ 'e' ], …(2) } to deeply equal { classification: [ 'e' ], …(2) } - Expected + Received @@ -15,14 +15,14 @@ 7, ], }, "title": [ { - "source": "Meet", + "source": "Meet", "span": [ 8, - 15, + 12, ], }, ], }, ], ❯ test/markdown-rows.test.ts:26:15 24| return value 25| })) 26| expect(got).toEqual(v.expected) | ^ 27| }) 28| for (const v of original.vectors) it(`candidate original line vector: … ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯[2/2]⎯ Test Files 1 failed | 2 passed (3) Tests 2 failed | 177 passed (179) Start at 19:53:43 Duration 1.34s (environment 50%, import 17%, transform 16%, tests 8%, setup 7%, worker 2%) $ tsc --noEmit GATE_STATUS native=101 seed=0 composer=0 install=0 editor=1 types=0 Composer byte comparison: recovered before == recovered after == fresh 2da88cbd: 12/12 Composer byte arrays identical composer "code-does-not-create-link": [96, 91, 80, 114, 105, 121, 97, 93, 40, 60, 112, 46, 109, 100, 62, 41, 96] composer "composer-code-in-label": [77, 101, 101, 116, 32, 97, 32, 98] composer "composer-emphasis-in-label": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "composer-formatting-outside-link": [42, 42, 77, 101, 101, 116, 42, 42, 32, 80, 114, 105, 121, 97, 32, 116, 111, 100, 97, 121] composer "composer-person-link": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "composer-strikethrough-compatibility": [77, 101, 101, 116, 32, 126, 126, 111, 108, 100, 126, 126] composer "composer-subscript-compatibility": [77, 101, 101, 116, 32, 72, 126, 50, 126, 79] composer "composer-superscript-compatibility": [77, 101, 101, 116, 32, 120, 94, 50, 94] composer "emoji-source-label": [77, 101, 101, 116, 32, 58, 117, 110, 107, 110, 111, 119, 110, 95, 78, 65, 77, 69, 58] composer "ordinary-source": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "two-links": [65, 32, 97, 110, 100, 32, 66] composer "unicode-byte-span": [240, 159, 152, 128, 32, 99, 97, 102, 195, 169] Corpus SHA256: b9d3635b6eb121610e5aefe54492144f526b38c1bf50902af111c96fb096632e contracts/markdown-extraction/composer.json 8346eaf880bbb8940bf642e27f473fe639caf1017e8b5254192fd44b1e9ca838 contracts/markdown-extraction/links.json 94afdf93c5f61b6821334324794512c5a62d5f1b52799d56c018200b0ae69c3a contracts/markdown-extraction/rows.json ```

kayg commented

2026-11-02 12:42:04 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-final-perf-evidence.txt

Source commit: 195168428eb02cced156c9247a6b226f4b68bce0; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033523,"p95_ms":0.047809000000000004,"p99_ms":0.055985},{"bytes":32768,"p50_ms":0.255359,"p95_ms":0.340017,"p99_ms":0.371197},{"bytes":262144,"p50_ms":2.5715489999999996,"p95_ms":2.987309,"p99_ms":3.381337}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.534855906}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033523, "p95_ms": 0.047809000000000004, "p99_ms": 0.055985}, {"bytes": 32768, "p50_ms": 0.255359, "p95_ms": 0.340017, "p99_ms": 0.371197}, {"bytes": 262144, "p50_ms": 2.5715489999999996, "p95_ms": 2.987309, "p99_ms": 3.381337}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.534855906, "cpu_s": 0.83689, "maxrss_kib": 20000}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.037751,"p95_ms":0.052519,"p99_ms":0.064621},{"bytes":32768,"p50_ms":0.267442,"p95_ms":0.357389,"p99_ms":0.404248},{"bytes":262144,"p50_ms":2.368118,"p95_ms":2.6433429999999998,"p99_ms":2.819624}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.550398772}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.037751, "p95_ms": 0.052519, "p99_ms": 0.064621}, {"bytes": 32768, "p50_ms": 0.267442, "p95_ms": 0.357389, "p99_ms": 0.404248}, {"bytes": 262144, "p50_ms": 2.368118, "p95_ms": 2.6433429999999998, "p99_ms": 2.819624}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.550398772, "cpu_s": 0.878979, "maxrss_kib": 20000}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033673,"p95_ms":0.048429,"p99_ms":0.057968},{"bytes":32768,"p50_ms":0.259716,"p95_ms":0.35236,"p99_ms":0.38402899999999995},{"bytes":262144,"p50_ms":2.596426,"p95_ms":3.253928,"p99_ms":3.599386}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.457216405}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033673, "p95_ms": 0.048429, "p99_ms": 0.057968}, {"bytes": 32768, "p50_ms": 0.259716, "p95_ms": 0.35236, "p99_ms": 0.38402899999999995}, {"bytes": 262144, "p50_ms": 2.596426, "p95_ms": 3.253928, "p99_ms": 3.599386}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.457216405, "cpu_s": 8.582658, "maxrss_kib": 20000}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.036649,"p95_ms":0.051967,"p99_ms":0.063459},{"bytes":32768,"p50_ms":0.257333,"p95_ms":0.344817,"p99_ms":0.388968},{"bytes":262144,"p50_ms":2.308707,"p95_ms":2.938627,"p99_ms":3.550814}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.52657725}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.036649, "p95_ms": 0.051967, "p99_ms": 0.063459}, {"bytes": 32768, "p50_ms": 0.257333, "p95_ms": 0.344817, "p99_ms": 0.388968}, {"bytes": 262144, "p50_ms": 2.308707, "p95_ms": 2.938627, "p99_ms": 3.550814}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.52657725, "cpu_s": 8.714830000000001, "maxrss_kib": 20000}
BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.0035379999999999995, "growth": 0.07305540068966931, "added_pass": true, "growth_pass": true}, {"bytes": 32768, "added_p95_ms": -0.007543000000000022, "growth": -0.021407083664434112, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.3153010000000003, "growth": -0.09689857919413103, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5618695695259236, "vs_paired_direct": 0.012709931190643342, "historical_wall_pass": true, "paired_wall_pass": true}
CONTAINMENT {"probe": "ordinary", "bytes": 24, "exit": 0, "stdout": "{\"elapsed_ms\":0.642915,\"outcome\":{\"rows\":2,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "depth255", "bytes": 518, "exit": 0, "stdout": "{\"elapsed_ms\":0.707606,\"outcome\":{\"rows\":1,\"status\":\"accepted\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "depth256", "bytes": 520, "exit": 0, "stdout": "{\"elapsed_ms\":0.042209,\"outcome\":{\"reason\":\"depth-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "long-paragraph", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":0.855203,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "short-line-storm", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":0.881032,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "inline-storm", "bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":0.454592,\"outcome\":{\"reason\":\"memory-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}
CONTAINMENT {"probe": "source-limit", "bytes": 8388608, "exit": 0, "stdout": "{\"elapsed_ms\":0.9145540000000001,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""}

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-final-perf-evidence.txt` Source commit: [195168428eb02cced156c9247a6b226f4b68bce0](https://git.kayg.org/kayg/calternal/commit/195168428eb02cced156c9247a6b226f4b68bce0); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033523,"p95_ms":0.047809000000000004,"p99_ms":0.055985},{"bytes":32768,"p50_ms":0.255359,"p95_ms":0.340017,"p99_ms":0.371197},{"bytes":262144,"p50_ms":2.5715489999999996,"p95_ms":2.987309,"p99_ms":3.381337}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.534855906} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033523, "p95_ms": 0.047809000000000004, "p99_ms": 0.055985}, {"bytes": 32768, "p50_ms": 0.255359, "p95_ms": 0.340017, "p99_ms": 0.371197}, {"bytes": 262144, "p50_ms": 2.5715489999999996, "p95_ms": 2.987309, "p99_ms": 3.381337}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.534855906, "cpu_s": 0.83689, "maxrss_kib": 20000} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.037751,"p95_ms":0.052519,"p99_ms":0.064621},{"bytes":32768,"p50_ms":0.267442,"p95_ms":0.357389,"p99_ms":0.404248},{"bytes":262144,"p50_ms":2.368118,"p95_ms":2.6433429999999998,"p99_ms":2.819624}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.550398772} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.037751, "p95_ms": 0.052519, "p99_ms": 0.064621}, {"bytes": 32768, "p50_ms": 0.267442, "p95_ms": 0.357389, "p99_ms": 0.404248}, {"bytes": 262144, "p50_ms": 2.368118, "p95_ms": 2.6433429999999998, "p99_ms": 2.819624}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.550398772, "cpu_s": 0.878979, "maxrss_kib": 20000} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033673,"p95_ms":0.048429,"p99_ms":0.057968},{"bytes":32768,"p50_ms":0.259716,"p95_ms":0.35236,"p99_ms":0.38402899999999995},{"bytes":262144,"p50_ms":2.596426,"p95_ms":3.253928,"p99_ms":3.599386}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.457216405} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033673, "p95_ms": 0.048429, "p99_ms": 0.057968}, {"bytes": 32768, "p50_ms": 0.259716, "p95_ms": 0.35236, "p99_ms": 0.38402899999999995}, {"bytes": 262144, "p50_ms": 2.596426, "p95_ms": 3.253928, "p99_ms": 3.599386}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.457216405, "cpu_s": 8.582658, "maxrss_kib": 20000} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.036649,"p95_ms":0.051967,"p99_ms":0.063459},{"bytes":32768,"p50_ms":0.257333,"p95_ms":0.344817,"p99_ms":0.388968},{"bytes":262144,"p50_ms":2.308707,"p95_ms":2.938627,"p99_ms":3.550814}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.52657725} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.036649, "p95_ms": 0.051967, "p99_ms": 0.063459}, {"bytes": 32768, "p50_ms": 0.257333, "p95_ms": 0.344817, "p99_ms": 0.388968}, {"bytes": 262144, "p50_ms": 2.308707, "p95_ms": 2.938627, "p99_ms": 3.550814}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.52657725, "cpu_s": 8.714830000000001, "maxrss_kib": 20000} BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.0035379999999999995, "growth": 0.07305540068966931, "added_pass": true, "growth_pass": true}, {"bytes": 32768, "added_p95_ms": -0.007543000000000022, "growth": -0.021407083664434112, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.3153010000000003, "growth": -0.09689857919413103, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5618695695259236, "vs_paired_direct": 0.012709931190643342, "historical_wall_pass": true, "paired_wall_pass": true} CONTAINMENT {"probe": "ordinary", "bytes": 24, "exit": 0, "stdout": "{\"elapsed_ms\":0.642915,\"outcome\":{\"rows\":2,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "depth255", "bytes": 518, "exit": 0, "stdout": "{\"elapsed_ms\":0.707606,\"outcome\":{\"rows\":1,\"status\":\"accepted\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "depth256", "bytes": 520, "exit": 0, "stdout": "{\"elapsed_ms\":0.042209,\"outcome\":{\"reason\":\"depth-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "long-paragraph", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":0.855203,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "short-line-storm", "bytes": 8388607, "exit": 0, "stdout": "{\"elapsed_ms\":0.881032,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "inline-storm", "bytes": 1048576, "exit": 0, "stdout": "{\"elapsed_ms\":0.454592,\"outcome\":{\"reason\":\"memory-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} CONTAINMENT {"probe": "source-limit", "bytes": 8388608, "exit": 0, "stdout": "{\"elapsed_ms\":0.9145540000000001,\"outcome\":{\"reason\":\"source-limit\",\"status\":\"unresolved\"}}\n", "stderr_tail": ""} ```

kayg commented

2026-11-02 12:42:05 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-initial-perf-evidence.txt

FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033513,"p95_ms":0.047118,"p99_ms":0.055945},{"bytes":32768,"p50_ms":0.257703,"p95_ms":0.338514,"p99_ms":0.370164},{"bytes":262144,"p50_ms":2.572381,"p95_ms":3.059684,"p99_ms":3.1362170000000003}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.534543913}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033513, "p95_ms": 0.047118, "p99_ms": 0.055945}, {"bytes": 32768, "p50_ms": 0.257703, "p95_ms": 0.338514, "p99_ms": 0.370164}, {"bytes": 262144, "p50_ms": 2.572381, "p95_ms": 3.059684, "p99_ms": 3.1362170000000003}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.534543913, "cpu_s": 0.836352, "maxrss_kib": 19988}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.047659,"p95_ms":0.067456,"p99_ms":0.078717},{"bytes":32768,"p50_ms":0.34834299999999996,"p95_ms":0.440115,"p99_ms":0.493965},{"bytes":262144,"p50_ms":3.1179430000000004,"p95_ms":3.629873,"p99_ms":3.866377}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.698393038}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.047659, "p95_ms": 0.067456, "p99_ms": 0.078717}, {"bytes": 32768, "p50_ms": 0.34834299999999996, "p95_ms": 0.440115, "p99_ms": 0.493965}, {"bytes": 262144, "p50_ms": 3.1179430000000004, "p95_ms": 3.629873, "p99_ms": 3.866377}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.698393038, "cpu_s": 1.11578, "maxrss_kib": 19988}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033784,"p95_ms":0.048311,"p99_ms":0.057698},{"bytes":32768,"p50_ms":0.258154,"p95_ms":0.35027600000000003,"p99_ms":0.384151},{"bytes":262144,"p50_ms":2.569155,"p95_ms":3.180791,"p99_ms":3.4976640000000003}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.423331811}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033784, "p95_ms": 0.048311, "p99_ms": 0.057698}, {"bytes": 32768, "p50_ms": 0.258154, "p95_ms": 0.35027600000000003, "p99_ms": 0.384151}, {"bytes": 262144, "p50_ms": 2.569155, "p95_ms": 3.180791, "p99_ms": 3.4976640000000003}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.423331811, "cpu_s": 8.517066, "maxrss_kib": 19988}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.04807,"p95_ms":0.068508,"p99_ms":0.07956},{"bytes":32768,"p50_ms":0.349625,"p95_ms":0.464821,"p99_ms":0.505478},{"bytes":262144,"p50_ms":3.140686,"p95_ms":3.931708,"p99_ms":4.685672}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":7.158565175}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.04807, "p95_ms": 0.068508, "p99_ms": 0.07956}, {"bytes": 32768, "p50_ms": 0.349625, "p95_ms": 0.464821, "p99_ms": 0.505478}, {"bytes": 262144, "p50_ms": 3.140686, "p95_ms": 3.931708, "p99_ms": 4.685672}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 7.158565175, "cpu_s": 11.407295, "maxrss_kib": 19988}
BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.020197, "growth": 0.41806213905735756, "added_pass": true, "growth_pass": false}, {"bytes": 32768, "added_p95_ms": 0.11454499999999995, "growth": 0.32701355502518004, "added_pass": true, "growth_pass": false}, {"bytes": 262144, "added_p95_ms": 0.7509169999999998, "growth": 0.23607869866331987, "added_pass": false, "growth_pass": false}], "vs_historical_12.614s": -0.43249047288726816, "vs_paired_direct": 0.3199570714962474, "historical_wall_pass": true, "paired_wall_pass": false}
Traceback (most recent call last):
  File "/home/claude/calendar805-r2/compare-optimized.py", line 72, in <module>
    assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets
           ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError: [{'bytes': 4096, 'added_p95_ms': 0.020197, 'growth': 0.41806213905735756, 'added_pass': True, 'growth_pass': False}, {'bytes': 32768, 'added_p95_ms': 0.11454499999999995, 'growth': 0.32701355502518004, 'added_pass': True, 'growth_pass': False}, {'bytes': 262144, 'added_p95_ms': 0.7509169999999998, 'growth': 0.23607869866331987, 'added_pass': False, 'growth_pass': False}]

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-initial-perf-evidence.txt` Source commit: [195168428eb02cced156c9247a6b226f4b68bce0](https://git.kayg.org/kayg/calternal/commit/195168428eb02cced156c9247a6b226f4b68bce0); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033513,"p95_ms":0.047118,"p99_ms":0.055945},{"bytes":32768,"p50_ms":0.257703,"p95_ms":0.338514,"p99_ms":0.370164},{"bytes":262144,"p50_ms":2.572381,"p95_ms":3.059684,"p99_ms":3.1362170000000003}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.534543913} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033513, "p95_ms": 0.047118, "p99_ms": 0.055945}, {"bytes": 32768, "p50_ms": 0.257703, "p95_ms": 0.338514, "p99_ms": 0.370164}, {"bytes": 262144, "p50_ms": 2.572381, "p95_ms": 3.059684, "p99_ms": 3.1362170000000003}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.534543913, "cpu_s": 0.836352, "maxrss_kib": 19988} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.047659,"p95_ms":0.067456,"p99_ms":0.078717},{"bytes":32768,"p50_ms":0.34834299999999996,"p95_ms":0.440115,"p99_ms":0.493965},{"bytes":262144,"p50_ms":3.1179430000000004,"p95_ms":3.629873,"p99_ms":3.866377}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.698393038} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.047659, "p95_ms": 0.067456, "p99_ms": 0.078717}, {"bytes": 32768, "p50_ms": 0.34834299999999996, "p95_ms": 0.440115, "p99_ms": 0.493965}, {"bytes": 262144, "p50_ms": 3.1179430000000004, "p95_ms": 3.629873, "p99_ms": 3.866377}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.698393038, "cpu_s": 1.11578, "maxrss_kib": 19988} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033784,"p95_ms":0.048311,"p99_ms":0.057698},{"bytes":32768,"p50_ms":0.258154,"p95_ms":0.35027600000000003,"p99_ms":0.384151},{"bytes":262144,"p50_ms":2.569155,"p95_ms":3.180791,"p99_ms":3.4976640000000003}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.423331811} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033784, "p95_ms": 0.048311, "p99_ms": 0.057698}, {"bytes": 32768, "p50_ms": 0.258154, "p95_ms": 0.35027600000000003, "p99_ms": 0.384151}, {"bytes": 262144, "p50_ms": 2.569155, "p95_ms": 3.180791, "p99_ms": 3.4976640000000003}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.423331811, "cpu_s": 8.517066, "maxrss_kib": 19988} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.04807,"p95_ms":0.068508,"p99_ms":0.07956},{"bytes":32768,"p50_ms":0.349625,"p95_ms":0.464821,"p99_ms":0.505478},{"bytes":262144,"p50_ms":3.140686,"p95_ms":3.931708,"p99_ms":4.685672}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":7.158565175} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.04807, "p95_ms": 0.068508, "p99_ms": 0.07956}, {"bytes": 32768, "p50_ms": 0.349625, "p95_ms": 0.464821, "p99_ms": 0.505478}, {"bytes": 262144, "p50_ms": 3.140686, "p95_ms": 3.931708, "p99_ms": 4.685672}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 7.158565175, "cpu_s": 11.407295, "maxrss_kib": 19988} BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.020197, "growth": 0.41806213905735756, "added_pass": true, "growth_pass": false}, {"bytes": 32768, "added_p95_ms": 0.11454499999999995, "growth": 0.32701355502518004, "added_pass": true, "growth_pass": false}, {"bytes": 262144, "added_p95_ms": 0.7509169999999998, "growth": 0.23607869866331987, "added_pass": false, "growth_pass": false}], "vs_historical_12.614s": -0.43249047288726816, "vs_paired_direct": 0.3199570714962474, "historical_wall_pass": true, "paired_wall_pass": false} Traceback (most recent call last): File "/home/claude/calendar805-r2/compare-optimized.py", line 72, in <module> assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AssertionError: [{'bytes': 4096, 'added_p95_ms': 0.020197, 'growth': 0.41806213905735756, 'added_pass': True, 'growth_pass': False}, {'bytes': 32768, 'added_p95_ms': 0.11454499999999995, 'growth': 0.32701355502518004, 'added_pass': True, 'growth_pass': False}, {'bytes': 262144, 'added_p95_ms': 0.7509169999999998, 'growth': 0.23607869866331987, 'added_pass': False, 'growth_pass': False}] ```

kayg commented

2026-11-02 12:42:05 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-inline-perf-evidence.txt

FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033432,"p95_ms":0.047549,"p99_ms":0.057528},{"bytes":32768,"p50_ms":0.25598,"p95_ms":0.336029,"p99_ms":0.37903},{"bytes":262144,"p50_ms":2.567341,"p95_ms":3.117112,"p99_ms":3.346592}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.535354325}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033432, "p95_ms": 0.047549, "p99_ms": 0.057528}, {"bytes": 32768, "p50_ms": 0.25598, "p95_ms": 0.336029, "p99_ms": 0.37903}, {"bytes": 262144, "p50_ms": 2.567341, "p95_ms": 3.117112, "p99_ms": 3.346592}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.535354325, "cpu_s": 0.839691, "maxrss_kib": 20044}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.03727,"p95_ms":0.051527,"p99_ms":0.060634},{"bytes":32768,"p50_ms":0.259215,"p95_ms":0.33033999999999997,"p99_ms":0.359344},{"bytes":262144,"p50_ms":2.416628,"p95_ms":2.813052,"p99_ms":2.934489}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.535893984}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.03727, "p95_ms": 0.051527, "p99_ms": 0.060634}, {"bytes": 32768, "p50_ms": 0.259215, "p95_ms": 0.33033999999999997, "p99_ms": 0.359344}, {"bytes": 262144, "p50_ms": 2.416628, "p95_ms": 2.813052, "p99_ms": 2.934489}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.535893984, "cpu_s": 0.854647, "maxrss_kib": 20044}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033683000000000005,"p95_ms":0.04833,"p99_ms":0.058489},{"bytes":32768,"p50_ms":0.259326,"p95_ms":0.35109799999999997,"p99_ms":0.38903},{"bytes":262144,"p50_ms":2.595633,"p95_ms":3.347383,"p99_ms":3.609393}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.498369154}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033683000000000005, "p95_ms": 0.04833, "p99_ms": 0.058489}, {"bytes": 32768, "p50_ms": 0.259326, "p95_ms": 0.35109799999999997, "p99_ms": 0.38903}, {"bytes": 262144, "p50_ms": 2.595633, "p95_ms": 3.347383, "p99_ms": 3.609393}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.498369154, "cpu_s": 8.604637, "maxrss_kib": 20044}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.037821,"p95_ms":0.053209,"p99_ms":0.064501},{"bytes":32768,"p50_ms":0.264256,"p95_ms":0.349255,"p99_ms":0.390052},{"bytes":262144,"p50_ms":2.474397,"p95_ms":3.1273310000000003,"p99_ms":3.391966}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.580797728}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.037821, "p95_ms": 0.053209, "p99_ms": 0.064501}, {"bytes": 32768, "p50_ms": 0.264256, "p95_ms": 0.349255, "p99_ms": 0.390052}, {"bytes": 262144, "p50_ms": 2.474397, "p95_ms": 3.1273310000000003, "p99_ms": 3.391966}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.580797728, "cpu_s": 8.854213000000001, "maxrss_kib": 20044}
BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.004879000000000001, "growth": 0.10095178977860542, "added_pass": true, "growth_pass": false}, {"bytes": 32768, "added_p95_ms": -0.0018429999999999835, "growth": -0.005249246649083661, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.22005199999999991, "growth": -0.06573851871745773, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5575711330267956, "vs_paired_direct": 0.01499145868371432, "historical_wall_pass": true, "paired_wall_pass": true}
Traceback (most recent call last):
  File "/home/claude/calendar805-r2/compare-optimized.py", line 72, in <module>
    assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets
           ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError: [{'bytes': 4096, 'added_p95_ms': 0.004879000000000001, 'growth': 0.10095178977860542, 'added_pass': True, 'growth_pass': False}, {'bytes': 32768, 'added_p95_ms': -0.0018429999999999835, 'growth': -0.005249246649083661, 'added_pass': True, 'growth_pass': True}, {'bytes': 262144, 'added_p95_ms': -0.22005199999999991, 'growth': -0.06573851871745773, 'added_pass': True, 'growth_pass': True}]

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-inline-perf-evidence.txt` Source commit: [195168428eb02cced156c9247a6b226f4b68bce0](https://git.kayg.org/kayg/calternal/commit/195168428eb02cced156c9247a6b226f4b68bce0); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033432,"p95_ms":0.047549,"p99_ms":0.057528},{"bytes":32768,"p50_ms":0.25598,"p95_ms":0.336029,"p99_ms":0.37903},{"bytes":262144,"p50_ms":2.567341,"p95_ms":3.117112,"p99_ms":3.346592}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.535354325} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033432, "p95_ms": 0.047549, "p99_ms": 0.057528}, {"bytes": 32768, "p50_ms": 0.25598, "p95_ms": 0.336029, "p99_ms": 0.37903}, {"bytes": 262144, "p50_ms": 2.567341, "p95_ms": 3.117112, "p99_ms": 3.346592}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.535354325, "cpu_s": 0.839691, "maxrss_kib": 20044} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.03727,"p95_ms":0.051527,"p99_ms":0.060634},{"bytes":32768,"p50_ms":0.259215,"p95_ms":0.33033999999999997,"p99_ms":0.359344},{"bytes":262144,"p50_ms":2.416628,"p95_ms":2.813052,"p99_ms":2.934489}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.535893984} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.03727, "p95_ms": 0.051527, "p99_ms": 0.060634}, {"bytes": 32768, "p50_ms": 0.259215, "p95_ms": 0.33033999999999997, "p99_ms": 0.359344}, {"bytes": 262144, "p50_ms": 2.416628, "p95_ms": 2.813052, "p99_ms": 2.934489}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.535893984, "cpu_s": 0.854647, "maxrss_kib": 20044} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033683000000000005,"p95_ms":0.04833,"p99_ms":0.058489},{"bytes":32768,"p50_ms":0.259326,"p95_ms":0.35109799999999997,"p99_ms":0.38903},{"bytes":262144,"p50_ms":2.595633,"p95_ms":3.347383,"p99_ms":3.609393}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.498369154} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033683000000000005, "p95_ms": 0.04833, "p99_ms": 0.058489}, {"bytes": 32768, "p50_ms": 0.259326, "p95_ms": 0.35109799999999997, "p99_ms": 0.38903}, {"bytes": 262144, "p50_ms": 2.595633, "p95_ms": 3.347383, "p99_ms": 3.609393}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.498369154, "cpu_s": 8.604637, "maxrss_kib": 20044} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.037821,"p95_ms":0.053209,"p99_ms":0.064501},{"bytes":32768,"p50_ms":0.264256,"p95_ms":0.349255,"p99_ms":0.390052},{"bytes":262144,"p50_ms":2.474397,"p95_ms":3.1273310000000003,"p99_ms":3.391966}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.580797728} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.037821, "p95_ms": 0.053209, "p99_ms": 0.064501}, {"bytes": 32768, "p50_ms": 0.264256, "p95_ms": 0.349255, "p99_ms": 0.390052}, {"bytes": 262144, "p50_ms": 2.474397, "p95_ms": 3.1273310000000003, "p99_ms": 3.391966}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.580797728, "cpu_s": 8.854213000000001, "maxrss_kib": 20044} BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.004879000000000001, "growth": 0.10095178977860542, "added_pass": true, "growth_pass": false}, {"bytes": 32768, "added_p95_ms": -0.0018429999999999835, "growth": -0.005249246649083661, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.22005199999999991, "growth": -0.06573851871745773, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5575711330267956, "vs_paired_direct": 0.01499145868371432, "historical_wall_pass": true, "paired_wall_pass": true} Traceback (most recent call last): File "/home/claude/calendar805-r2/compare-optimized.py", line 72, in <module> assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AssertionError: [{'bytes': 4096, 'added_p95_ms': 0.004879000000000001, 'growth': 0.10095178977860542, 'added_pass': True, 'growth_pass': False}, {'bytes': 32768, 'added_p95_ms': -0.0018429999999999835, 'growth': -0.005249246649083661, 'added_pass': True, 'growth_pass': True}, {'bytes': 262144, 'added_p95_ms': -0.22005199999999991, 'growth': -0.06573851871745773, 'added_pass': True, 'growth_pass': True}] ```

kayg commented

2026-11-02 12:42:05 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-memchr-perf-evidence.txt

FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033833,"p95_ms":0.049613000000000004,"p99_ms":0.059251},{"bytes":32768,"p50_ms":0.260257,"p95_ms":0.35381300000000004,"p99_ms":0.383969},{"bytes":262144,"p50_ms":2.60922,"p95_ms":3.305685,"p99_ms":3.549481}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.545794147}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033833, "p95_ms": 0.049613000000000004, "p99_ms": 0.059251}, {"bytes": 32768, "p50_ms": 0.260257, "p95_ms": 0.35381300000000004, "p99_ms": 0.383969}, {"bytes": 262144, "p50_ms": 2.60922, "p95_ms": 3.305685, "p99_ms": 3.549481}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.545794147, "cpu_s": 0.857015, "maxrss_kib": 20036}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.038241000000000004,"p95_ms":0.052728000000000004,"p99_ms":0.060844},{"bytes":32768,"p50_ms":0.268774,"p95_ms":0.33735299999999996,"p99_ms":0.37141700000000005},{"bytes":262144,"p50_ms":2.387355,"p95_ms":2.81744,"p99_ms":2.9459310000000003}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.555905146}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.038241000000000004, "p95_ms": 0.052728000000000004, "p99_ms": 0.060844}, {"bytes": 32768, "p50_ms": 0.268774, "p95_ms": 0.33735299999999996, "p99_ms": 0.37141700000000005}, {"bytes": 262144, "p50_ms": 2.387355, "p95_ms": 2.81744, "p99_ms": 2.9459310000000003}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.555905146, "cpu_s": 0.881177, "maxrss_kib": 20036}
RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033643000000000006,"p95_ms":0.048221,"p99_ms":0.057117},{"bytes":32768,"p50_ms":0.258024,"p95_ms":0.34708100000000003,"p99_ms":0.37866},{"bytes":262144,"p50_ms":2.560578,"p95_ms":3.15356,"p99_ms":3.447091}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.410001576}
MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033643000000000006, "p95_ms": 0.048221, "p99_ms": 0.057117}, {"bytes": 32768, "p50_ms": 0.258024, "p95_ms": 0.34708100000000003, "p99_ms": 0.37866}, {"bytes": 262144, "p50_ms": 2.560578, "p95_ms": 3.15356, "p99_ms": 3.447091}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.410001576, "cpu_s": 8.478064, "maxrss_kib": 20036}
RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.038422,"p95_ms":0.053831,"p99_ms":0.063389},{"bytes":32768,"p50_ms":0.270137,"p95_ms":0.360695,"p99_ms":0.4007},{"bytes":262144,"p50_ms":2.474257,"p95_ms":3.023917,"p99_ms":3.342965}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.56377786}
MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.038422, "p95_ms": 0.053831, "p99_ms": 0.063389}, {"bytes": 32768, "p50_ms": 0.270137, "p95_ms": 0.360695, "p99_ms": 0.4007}, {"bytes": 262144, "p50_ms": 2.474257, "p95_ms": 3.023917, "p99_ms": 3.342965}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.56377786, "cpu_s": 8.941671, "maxrss_kib": 20036}
BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.005609999999999997, "growth": 0.11633935422326358, "added_pass": true, "growth_pass": false}, {"bytes": 32768, "added_p95_ms": 0.01361399999999996, "growth": 0.039224273296434964, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.12964300000000017, "growth": -0.0411100470579282, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5589204169969875, "vs_paired_direct": 0.028424443475614902, "historical_wall_pass": true, "paired_wall_pass": true}
Traceback (most recent call last):
  File "/home/claude/calendar805-r2/compare-optimized.py", line 72, in <module>
    assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets
           ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError: [{'bytes': 4096, 'added_p95_ms': 0.005609999999999997, 'growth': 0.11633935422326358, 'added_pass': True, 'growth_pass': False}, {'bytes': 32768, 'added_p95_ms': 0.01361399999999996, 'growth': 0.039224273296434964, 'added_pass': True, 'growth_pass': True}, {'bytes': 262144, 'added_p95_ms': -0.12964300000000017, 'growth': -0.0411100470579282, 'added_pass': True, 'growth_pass': True}]

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-memchr-perf-evidence.txt` Source commit: [195168428eb02cced156c9247a6b226f4b68bce0](https://git.kayg.org/kayg/calternal/commit/195168428eb02cced156c9247a6b226f4b68bce0); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text FIXTURE {"byte_identity_100_variants": true, "shard_sha256": ["75beb57f90105f254eaad4a289f2f3f392304ca5874eb773db722e5482b24c29", "15b9ceed9e51c07f602fdec18486a4842a17d16a471f3eaa345f18be50cbe69f"]} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033833,"p95_ms":0.049613000000000004,"p99_ms":0.059251},{"bytes":32768,"p50_ms":0.260257,"p95_ms":0.35381300000000004,"p99_ms":0.383969},{"bytes":262144,"p50_ms":2.60922,"p95_ms":3.305685,"p99_ms":3.549481}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"baseline","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.545794147} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033833, "p95_ms": 0.049613000000000004, "p99_ms": 0.059251}, {"bytes": 32768, "p50_ms": 0.260257, "p95_ms": 0.35381300000000004, "p99_ms": 0.383969}, {"bytes": 262144, "p50_ms": 2.60922, "p95_ms": 3.305685, "p99_ms": 3.549481}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "baseline", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.545794147, "cpu_s": 0.857015, "maxrss_kib": 20036} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.038241000000000004,"p95_ms":0.052728000000000004,"p99_ms":0.060844},{"bytes":32768,"p50_ms":0.268774,"p95_ms":0.33735299999999996,"p99_ms":0.37141700000000005},{"bytes":262144,"p50_ms":2.387355,"p95_ms":2.81744,"p99_ms":2.9459310000000003}],"bytes":92569600,"edges":5000,"isolated":0,"mode":"A","notes":10000,"rows":20000,"unresolved":0,"wall_s":0.555905146} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.038241000000000004, "p95_ms": 0.052728000000000004, "p99_ms": 0.060844}, {"bytes": 32768, "p50_ms": 0.268774, "p95_ms": 0.33735299999999996, "p99_ms": 0.37141700000000005}, {"bytes": 262144, "p50_ms": 2.387355, "p95_ms": 2.81744, "p99_ms": 2.9459310000000003}], "bytes": 92569600, "edges": 5000, "isolated": 0, "mode": "A", "notes": 10000, "rows": 20000, "unresolved": 0, "wall_s": 0.555905146, "cpu_s": 0.881177, "maxrss_kib": 20036} RAW options-baseline {"buckets":[{"bytes":4096,"p50_ms":0.033643000000000006,"p95_ms":0.048221,"p99_ms":0.057117},{"bytes":32768,"p50_ms":0.258024,"p95_ms":0.34708100000000003,"p99_ms":0.37866},{"bytes":262144,"p50_ms":2.560578,"p95_ms":3.15356,"p99_ms":3.447091}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"baseline","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.410001576} MEASURE options-baseline {"buckets": [{"bytes": 4096, "p50_ms": 0.033643000000000006, "p95_ms": 0.048221, "p99_ms": 0.057117}, {"bytes": 32768, "p50_ms": 0.258024, "p95_ms": 0.34708100000000003, "p99_ms": 0.37866}, {"bytes": 262144, "p50_ms": 2.560578, "p95_ms": 3.15356, "p99_ms": 3.447091}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "baseline", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.410001576, "cpu_s": 8.478064, "maxrss_kib": 20036} RAW options-head {"buckets":[{"bytes":4096,"p50_ms":0.038422,"p95_ms":0.053831,"p99_ms":0.063389},{"bytes":32768,"p50_ms":0.270137,"p95_ms":0.360695,"p99_ms":0.4007},{"bytes":262144,"p50_ms":2.474257,"p95_ms":3.023917,"p99_ms":3.342965}],"bytes":925696000,"edges":50000,"isolated":0,"mode":"A","notes":100000,"rows":200000,"unresolved":0,"wall_s":5.56377786} MEASURE options-head {"buckets": [{"bytes": 4096, "p50_ms": 0.038422, "p95_ms": 0.053831, "p99_ms": 0.063389}, {"bytes": 32768, "p50_ms": 0.270137, "p95_ms": 0.360695, "p99_ms": 0.4007}, {"bytes": 262144, "p50_ms": 2.474257, "p95_ms": 3.023917, "p99_ms": 3.342965}], "bytes": 925696000, "edges": 50000, "isolated": 0, "mode": "A", "notes": 100000, "rows": 200000, "unresolved": 0, "wall_s": 5.56377786, "cpu_s": 8.941671, "maxrss_kib": 20036} BUDGET {"option": "A", "buckets": [{"bytes": 4096, "added_p95_ms": 0.005609999999999997, "growth": 0.11633935422326358, "added_pass": true, "growth_pass": false}, {"bytes": 32768, "added_p95_ms": 0.01361399999999996, "growth": 0.039224273296434964, "added_pass": true, "growth_pass": true}, {"bytes": 262144, "added_p95_ms": -0.12964300000000017, "growth": -0.0411100470579282, "added_pass": true, "growth_pass": true}], "vs_historical_12.614s": -0.5589204169969875, "vs_paired_direct": 0.028424443475614902, "historical_wall_pass": true, "paired_wall_pass": true} Traceback (most recent call last): File "/home/claude/calendar805-r2/compare-optimized.py", line 72, in <module> assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AssertionError: [{'bytes': 4096, 'added_p95_ms': 0.005609999999999997, 'growth': 0.11633935422326358, 'added_pass': True, 'growth_pass': False}, {'bytes': 32768, 'added_p95_ms': 0.01361399999999996, 'growth': 0.039224273296434964, 'added_pass': True, 'growth_pass': True}, {'bytes': 262144, 'added_p95_ms': -0.12964300000000017, 'growth': -0.0411100470579282, 'added_pass': True, 'growth_pass': True}] ```

kayg commented

2026-11-02 12:42:06 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-memchr-profile-evidence.txt

Source commit: 948117ccc9dbcfa02e044dc5efc6484a32f414e1; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
==116510== Callgrind, a call-graph generating cache profiler
==116510== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al.
==116510== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info
==116510== Command: /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 12 1000
==116510==
==116510== For interactive control, run 'callgrind_control -h'.
==116510==
==116510== Events    : Ir
==116510== Collected : 439313911
==116510==
==116510== I   refs:      439,313,911
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
	LANGUAGE = (unset),
	LC_ALL = (unset),
	LC_CTYPE = (unset),
	LC_NUMERIC = (unset),
	LC_COLLATE = (unset),
	LC_TIME = (unset),
	LC_MESSAGES = (unset),
	LC_MONETARY = (unset),
	LC_ADDRESS = (unset),
	LC_IDENTIFICATION = (unset),
	LC_MEASUREMENT = (unset),
	LC_PAPER = (unset),
	LC_TELEPHONE = (unset),
	LC_NAME = (unset),
	LANG = "en_US.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
--------------------------------------------------------------------------------
Profile data file '/home/claude/calendar805-r2-fast.callgrind' (creator: callgrind-3.27.1)
--------------------------------------------------------------------------------
I1 cache:
D1 cache:
LL cache:
Timerange: Basic block 0 - 93210309
Trigger: Program termination
Profiled target:  /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 12 1000 (PID 116510, part 1)
Events recorded:  Ir
Events shown:     Ir
Event sort order: Ir
Thresholds:       85
Include dirs:
User annotated:
Auto-annotation:  on

--------------------------------------------------------------------------------
Ir
--------------------------------------------------------------------------------
439,313,911 (100.0%)  PROGRAM TOTALS

--------------------------------------------------------------------------------
Ir                   file:function
--------------------------------------------------------------------------------
74,284,210 (16.91%)  ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
38,250,212 ( 8.71%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
33,323,290 ( 7.59%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
32,509,477 ( 7.40%)  ???:<pulldown_cmark::parse::OffsetIter as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
28,028,000 ( 6.38%)  ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
19,741,722 ( 4.49%)  ???:calternal_markdown_view::source::lines::{closure#0} [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
18,451,433 ( 4.20%)  ???:<calternal_markdown_view::bounded::Budget>::new [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
17,206,326 ( 3.92%)  ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
14,998,992 ( 3.41%)  ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
13,905,892 ( 3.17%)  ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
13,059,046 ( 2.97%)  ???:<u8 as core::slice::cmp::SliceContains>::slice_contains [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
12,833,821 ( 2.92%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
10,500,490 ( 2.39%)  ???:memchr::arch::x86_64::memchr::memchr2_raw::find_avx2 [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 9,069,060 ( 2.06%)  ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 7,976,969 ( 1.82%)  ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 7,805,798 ( 1.78%)  ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 7,281,274 ( 1.66%)  ???:<calternal_markdown_view::bounded::Budget>::finish::<calternal_markdown_view::lines::Extraction> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 6,576,570 ( 1.50%)  ???:calternal_markdown_view::lines::extract_with_limits::{closure#2} [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 5,200,428 ( 1.18%)  ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
 5,109,104 ( 1.16%)  ???:<regex_automata::hybrid::regex::Regex>::try_search [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-memchr-profile-evidence.txt` Source commit: [948117ccc9dbcfa02e044dc5efc6484a32f414e1](https://git.kayg.org/kayg/calternal/commit/948117ccc9dbcfa02e044dc5efc6484a32f414e1); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead ==116510== Callgrind, a call-graph generating cache profiler ==116510== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al. ==116510== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info ==116510== Command: /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 12 1000 ==116510== ==116510== For interactive control, run 'callgrind_control -h'. ==116510== ==116510== Events : Ir ==116510== Collected : 439313911 ==116510== ==116510== I refs: 439,313,911 perl: warning: Setting locale failed. perl: warning: Please check that your locale settings: LANGUAGE = (unset), LC_ALL = (unset), LC_CTYPE = (unset), LC_NUMERIC = (unset), LC_COLLATE = (unset), LC_TIME = (unset), LC_MESSAGES = (unset), LC_MONETARY = (unset), LC_ADDRESS = (unset), LC_IDENTIFICATION = (unset), LC_MEASUREMENT = (unset), LC_PAPER = (unset), LC_TELEPHONE = (unset), LC_NAME = (unset), LANG = "en_US.UTF-8" are supported and installed on your system. perl: warning: Falling back to the standard locale ("C"). -------------------------------------------------------------------------------- Profile data file '/home/claude/calendar805-r2-fast.callgrind' (creator: callgrind-3.27.1) -------------------------------------------------------------------------------- I1 cache: D1 cache: LL cache: Timerange: Basic block 0 - 93210309 Trigger: Program termination Profiled target: /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 12 1000 (PID 116510, part 1) Events recorded: Ir Events shown: Ir Event sort order: Ir Thresholds: 85 Include dirs: User annotated: Auto-annotation: on -------------------------------------------------------------------------------- Ir -------------------------------------------------------------------------------- 439,313,911 (100.0%) PROGRAM TOTALS -------------------------------------------------------------------------------- Ir file:function -------------------------------------------------------------------------------- 74,284,210 (16.91%) ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 38,250,212 ( 8.71%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 33,323,290 ( 7.59%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 32,509,477 ( 7.40%) ???:<pulldown_cmark::parse::OffsetIter as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 28,028,000 ( 6.38%) ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 19,741,722 ( 4.49%) ???:calternal_markdown_view::source::lines::{closure#0} [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 18,451,433 ( 4.20%) ???:<calternal_markdown_view::bounded::Budget>::new [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 17,206,326 ( 3.92%) ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 14,998,992 ( 3.41%) ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 13,905,892 ( 3.17%) ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 13,059,046 ( 2.97%) ???:<u8 as core::slice::cmp::SliceContains>::slice_contains [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 12,833,821 ( 2.92%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 10,500,490 ( 2.39%) ???:memchr::arch::x86_64::memchr::memchr2_raw::find_avx2 [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 9,069,060 ( 2.06%) ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 7,976,969 ( 1.82%) ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 7,805,798 ( 1.78%) ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 7,281,274 ( 1.66%) ???:<calternal_markdown_view::bounded::Budget>::finish::<calternal_markdown_view::lines::Extraction> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 6,576,570 ( 1.50%) ???:calternal_markdown_view::lines::extract_with_limits::{closure#2} [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 5,200,428 ( 1.18%) ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 5,109,104 ( 1.16%) ???:<regex_automata::hybrid::regex::Regex>::try_search [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] ```

kayg commented

2026-11-02 12:42:06 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-memory-evidence.jsonl

{"measured_probes": 136, "peak_rss_kib": 30052, "limit_as_mib": 256, "limit_stack_mib": 8, "rss_guard_mib": 32}
{"shape": "bare-cr", "bytes": 4194304, "reason": "line-limit", "status": "unresolved"}
{"shape": "bare-cr", "bytes": 1048576, "reason": "line-limit", "status": "unresolved"}
{"bytes": 4194304, "reason": "blank-run-limit", "status": "unresolved"}
{"bytes": 1048576, "reason": "blank-run-limit", "status": "unresolved"}
{"ending": "b'\\r'", "bytes": 1048576, "reason": "line-limit", "status": "unresolved"}
{"ending": "b'\\r'", "bytes": 4194304, "reason": "line-limit", "status": "unresolved"}
{"ending": "b'\\n'", "bytes": 1048576, "reason": "line-limit", "status": "unresolved"}
{"ending": "b'\\n'", "bytes": 4194304, "reason": "line-limit", "status": "unresolved"}
{"ending": "b'\\r\\n'", "bytes": 1048575, "reason": "line-limit", "status": "unresolved"}
{"ending": "b'\\r\\n'", "bytes": 4194303, "reason": "line-limit", "status": "unresolved"}
{"ending": "b'\\r\\n\\n\\r'", "bytes": 1048575, "reason": "line-limit", "status": "unresolved"}
{"ending": "b'\\r\\n\\n\\r'", "bytes": 4194300, "reason": "line-limit", "status": "unresolved"}
{"shape": "long-event-line-4MiB", "bytes": 4194304, "maxrss_kib": 23840, "elapsed_ms": 8.616137, "outcome": {"rows": 1, "status": "accepted"}}
{"shape": "blank-lines-4MiB", "bytes": 4194304, "maxrss_kib": 25120, "elapsed_ms": 18.359816000000002, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "deep-lists-4MiB", "bytes": 4194304, "maxrss_kib": 29260, "elapsed_ms": 9.516013000000001, "outcome": {"rows": 1, "status": "accepted"}}
{"shape": "tags", "max_accepted_tokens": 98428, "first_refused_tokens": 98429, "sweep_peak_kib": 29268}
{"shape": "links", "max_accepted_tokens": 39870, "first_refused_tokens": 39871, "sweep_peak_kib": 29272}
{"property": "mixed-endings-blank-runs-depth", "seed": 80502, "cases": 72, "accepted": 37, "refused": 35}
{"shape": "line-edge-b'\\r'-16383", "bytes": 32765, "maxrss_kib": 30052, "elapsed_ms": 3.9910889999999997, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\r'-16384", "bytes": 32767, "maxrss_kib": 30052, "elapsed_ms": 4.246839, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\r'-16385", "bytes": 32769, "maxrss_kib": 30052, "elapsed_ms": 0.172724, "outcome": {"reason": "line-limit", "status": "unresolved"}}
{"shape": "blank-edge-b'\\r'-1023", "bytes": 2045, "maxrss_kib": 30052, "elapsed_ms": 0.41586900000000004, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "blank-edge-b'\\r'-1024", "bytes": 2047, "maxrss_kib": 30052, "elapsed_ms": 0.393236, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "blank-edge-b'\\r'-1025", "bytes": 2049, "maxrss_kib": 30052, "elapsed_ms": 0.019506, "outcome": {"reason": "blank-run-limit", "status": "unresolved"}}
{"shape": "deep-b'\\r'", "bytes": 4194304, "maxrss_kib": 30052, "elapsed_ms": 5.807653999999999, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\n'-16383", "bytes": 32765, "maxrss_kib": 30052, "elapsed_ms": 4.21017, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\n'-16384", "bytes": 32767, "maxrss_kib": 30052, "elapsed_ms": 4.099450999999999, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\n'-16385", "bytes": 32769, "maxrss_kib": 30052, "elapsed_ms": 0.131968, "outcome": {"reason": "line-limit", "status": "unresolved"}}
{"shape": "blank-edge-b'\\n'-1023", "bytes": 2045, "maxrss_kib": 30052, "elapsed_ms": 0.43069799999999997, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "blank-edge-b'\\n'-1024", "bytes": 2047, "maxrss_kib": 30052, "elapsed_ms": 0.393148, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "blank-edge-b'\\n'-1025", "bytes": 2049, "maxrss_kib": 30052, "elapsed_ms": 0.022001000000000003, "outcome": {"reason": "blank-run-limit", "status": "unresolved"}}
{"shape": "deep-b'\\n'", "bytes": 4194304, "maxrss_kib": 30052, "elapsed_ms": 5.593152, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\r\\n'-16383", "bytes": 49147, "maxrss_kib": 30052, "elapsed_ms": 5.290825, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\r\\n'-16384", "bytes": 49150, "maxrss_kib": 30052, "elapsed_ms": 5.322875000000001, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "line-edge-b'\\r\\n'-16385", "bytes": 49153, "maxrss_kib": 30052, "elapsed_ms": 0.17582, "outcome": {"reason": "line-limit", "status": "unresolved"}}
{"shape": "blank-edge-b'\\r\\n'-1023", "bytes": 3067, "maxrss_kib": 30052, "elapsed_ms": 0.484798, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "blank-edge-b'\\r\\n'-1024", "bytes": 3070, "maxrss_kib": 30052, "elapsed_ms": 0.439513, "outcome": {"rows": 0, "status": "accepted"}}
{"shape": "blank-edge-b'\\r\\n'-1025", "bytes": 3073, "maxrss_kib": 30052, "elapsed_ms": 0.020378, "outcome": {"reason": "blank-run-limit", "status": "unresolved"}}
{"shape": "deep-b'\\r\\n'", "bytes": 4194304, "maxrss_kib": 30052, "elapsed_ms": 6.261465, "outcome": {"rows": 0, "status": "accepted"}}

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-memory-evidence.jsonl` Source commit: [195168428eb02cced156c9247a6b226f4b68bce0](https://git.kayg.org/kayg/calternal/commit/195168428eb02cced156c9247a6b226f4b68bce0); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text {"measured_probes": 136, "peak_rss_kib": 30052, "limit_as_mib": 256, "limit_stack_mib": 8, "rss_guard_mib": 32} {"shape": "bare-cr", "bytes": 4194304, "reason": "line-limit", "status": "unresolved"} {"shape": "bare-cr", "bytes": 1048576, "reason": "line-limit", "status": "unresolved"} {"bytes": 4194304, "reason": "blank-run-limit", "status": "unresolved"} {"bytes": 1048576, "reason": "blank-run-limit", "status": "unresolved"} {"ending": "b'\\r'", "bytes": 1048576, "reason": "line-limit", "status": "unresolved"} {"ending": "b'\\r'", "bytes": 4194304, "reason": "line-limit", "status": "unresolved"} {"ending": "b'\\n'", "bytes": 1048576, "reason": "line-limit", "status": "unresolved"} {"ending": "b'\\n'", "bytes": 4194304, "reason": "line-limit", "status": "unresolved"} {"ending": "b'\\r\\n'", "bytes": 1048575, "reason": "line-limit", "status": "unresolved"} {"ending": "b'\\r\\n'", "bytes": 4194303, "reason": "line-limit", "status": "unresolved"} {"ending": "b'\\r\\n\\n\\r'", "bytes": 1048575, "reason": "line-limit", "status": "unresolved"} {"ending": "b'\\r\\n\\n\\r'", "bytes": 4194300, "reason": "line-limit", "status": "unresolved"} {"shape": "long-event-line-4MiB", "bytes": 4194304, "maxrss_kib": 23840, "elapsed_ms": 8.616137, "outcome": {"rows": 1, "status": "accepted"}} {"shape": "blank-lines-4MiB", "bytes": 4194304, "maxrss_kib": 25120, "elapsed_ms": 18.359816000000002, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "deep-lists-4MiB", "bytes": 4194304, "maxrss_kib": 29260, "elapsed_ms": 9.516013000000001, "outcome": {"rows": 1, "status": "accepted"}} {"shape": "tags", "max_accepted_tokens": 98428, "first_refused_tokens": 98429, "sweep_peak_kib": 29268} {"shape": "links", "max_accepted_tokens": 39870, "first_refused_tokens": 39871, "sweep_peak_kib": 29272} {"property": "mixed-endings-blank-runs-depth", "seed": 80502, "cases": 72, "accepted": 37, "refused": 35} {"shape": "line-edge-b'\\r'-16383", "bytes": 32765, "maxrss_kib": 30052, "elapsed_ms": 3.9910889999999997, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\r'-16384", "bytes": 32767, "maxrss_kib": 30052, "elapsed_ms": 4.246839, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\r'-16385", "bytes": 32769, "maxrss_kib": 30052, "elapsed_ms": 0.172724, "outcome": {"reason": "line-limit", "status": "unresolved"}} {"shape": "blank-edge-b'\\r'-1023", "bytes": 2045, "maxrss_kib": 30052, "elapsed_ms": 0.41586900000000004, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "blank-edge-b'\\r'-1024", "bytes": 2047, "maxrss_kib": 30052, "elapsed_ms": 0.393236, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "blank-edge-b'\\r'-1025", "bytes": 2049, "maxrss_kib": 30052, "elapsed_ms": 0.019506, "outcome": {"reason": "blank-run-limit", "status": "unresolved"}} {"shape": "deep-b'\\r'", "bytes": 4194304, "maxrss_kib": 30052, "elapsed_ms": 5.807653999999999, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\n'-16383", "bytes": 32765, "maxrss_kib": 30052, "elapsed_ms": 4.21017, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\n'-16384", "bytes": 32767, "maxrss_kib": 30052, "elapsed_ms": 4.099450999999999, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\n'-16385", "bytes": 32769, "maxrss_kib": 30052, "elapsed_ms": 0.131968, "outcome": {"reason": "line-limit", "status": "unresolved"}} {"shape": "blank-edge-b'\\n'-1023", "bytes": 2045, "maxrss_kib": 30052, "elapsed_ms": 0.43069799999999997, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "blank-edge-b'\\n'-1024", "bytes": 2047, "maxrss_kib": 30052, "elapsed_ms": 0.393148, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "blank-edge-b'\\n'-1025", "bytes": 2049, "maxrss_kib": 30052, "elapsed_ms": 0.022001000000000003, "outcome": {"reason": "blank-run-limit", "status": "unresolved"}} {"shape": "deep-b'\\n'", "bytes": 4194304, "maxrss_kib": 30052, "elapsed_ms": 5.593152, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\r\\n'-16383", "bytes": 49147, "maxrss_kib": 30052, "elapsed_ms": 5.290825, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\r\\n'-16384", "bytes": 49150, "maxrss_kib": 30052, "elapsed_ms": 5.322875000000001, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "line-edge-b'\\r\\n'-16385", "bytes": 49153, "maxrss_kib": 30052, "elapsed_ms": 0.17582, "outcome": {"reason": "line-limit", "status": "unresolved"}} {"shape": "blank-edge-b'\\r\\n'-1023", "bytes": 3067, "maxrss_kib": 30052, "elapsed_ms": 0.484798, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "blank-edge-b'\\r\\n'-1024", "bytes": 3070, "maxrss_kib": 30052, "elapsed_ms": 0.439513, "outcome": {"rows": 0, "status": "accepted"}} {"shape": "blank-edge-b'\\r\\n'-1025", "bytes": 3073, "maxrss_kib": 30052, "elapsed_ms": 0.020378, "outcome": {"reason": "blank-run-limit", "status": "unresolved"}} {"shape": "deep-b'\\r\\n'", "bytes": 4194304, "maxrss_kib": 30052, "elapsed_ms": 6.261465, "outcome": {"rows": 0, "status": "accepted"}} ```

kayg commented

2026-11-02 12:42:06 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-profile-evidence.txt

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
==101436== Callgrind, a call-graph generating cache profiler
==101436== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al.
==101436== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info
==101436== Command: /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 90 100
==101436==
==101436== For interactive control, run 'callgrind_control -h'.
==101436==
==101436== Events    : Ir
==101436== Collected : 412815570
==101436==
==101436== I   refs:      412,815,570
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
	LANGUAGE = (unset),
	LC_ALL = (unset),
	LC_CTYPE = (unset),
	LC_NUMERIC = (unset),
	LC_COLLATE = (unset),
	LC_TIME = (unset),
	LC_MESSAGES = (unset),
	LC_MONETARY = (unset),
	LC_ADDRESS = (unset),
	LC_IDENTIFICATION = (unset),
	LC_MEASUREMENT = (unset),
	LC_PAPER = (unset),
	LC_TELEPHONE = (unset),
	LC_NAME = (unset),
	LANG = "en_US.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
--------------------------------------------------------------------------------
Profile data file '/home/claude/calendar805-r2.callgrind' (creator: callgrind-3.27.1)
--------------------------------------------------------------------------------
I1 cache:
D1 cache:
LL cache:
Timerange: Basic block 0 - 99513946
Trigger: Program termination
Profiled target:  /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 90 100 (PID 101436, part 1)
Events recorded:  Ir
Events shown:     Ir
Event sort order: Ir
Thresholds:       90
Include dirs:
User annotated:
Auto-annotation:  on

--------------------------------------------------------------------------------
Ir
--------------------------------------------------------------------------------
412,815,570 (100.0%)  PROGRAM TOTALS

--------------------------------------------------------------------------------
Ir                    file:function
--------------------------------------------------------------------------------
113,484,610 (27.49%)  ???:calternal_markdown_view::source::lines::{closure#0} [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 58,447,185 (14.16%)  ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 29,856,206 ( 7.23%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 26,222,428 ( 6.35%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 22,170,813 ( 5.37%)  ???:<pulldown_cmark::parse::OffsetIter as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 16,560,061 ( 4.01%)  ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 14,602,883 ( 3.54%)  ???:<calternal_markdown_view::bounded::Budget>::new [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 13,026,906 ( 3.16%)  ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 11,912,352 ( 2.89%)  ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 10,966,075 ( 2.66%)  ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 10,716,504 ( 2.60%)  ???:<u8 as core::slice::cmp::SliceContains>::slice_contains [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
 10,081,719 ( 2.44%)  ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
  6,512,343 ( 1.58%)  ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
  6,187,462 ( 1.50%)  ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
  5,909,813 ( 1.43%)  ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
  5,511,484 ( 1.34%)  ???:__memcpy_avx_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6]
  5,479,856 ( 1.33%)  ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]
  5,344,011 ( 1.29%)  ???:<calternal_markdown_view::bounded::Budget>::finish::<calternal_markdown_view::lines::Extraction> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options]

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-profile-evidence.txt` Source commit: [948117ccc9dbcfa02e044dc5efc6484a32f414e1](https://git.kayg.org/kayg/calternal/commit/948117ccc9dbcfa02e044dc5efc6484a32f414e1); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead ==101436== Callgrind, a call-graph generating cache profiler ==101436== Copyright (C) 2002-2017, and GNU GPL'd, by Josef Weidendorfer et al. ==101436== Using Valgrind-3.27.1 and LibVEX; rerun with -h for copyright info ==101436== Command: /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 90 100 ==101436== ==101436== For interactive control, run 'callgrind_control -h'. ==101436== ==101436== Events : Ir ==101436== Collected : 412815570 ==101436== ==101436== I refs: 412,815,570 perl: warning: Setting locale failed. perl: warning: Please check that your locale settings: LANGUAGE = (unset), LC_ALL = (unset), LC_CTYPE = (unset), LC_NUMERIC = (unset), LC_COLLATE = (unset), LC_TIME = (unset), LC_MESSAGES = (unset), LC_MONETARY = (unset), LC_ADDRESS = (unset), LC_IDENTIFICATION = (unset), LC_MEASUREMENT = (unset), LC_PAPER = (unset), LC_TELEPHONE = (unset), LC_NAME = (unset), LANG = "en_US.UTF-8" are supported and installed on your system. perl: warning: Falling back to the standard locale ("C"). -------------------------------------------------------------------------------- Profile data file '/home/claude/calendar805-r2.callgrind' (creator: callgrind-3.27.1) -------------------------------------------------------------------------------- I1 cache: D1 cache: LL cache: Timerange: Basic block 0 - 99513946 Trigger: Program termination Profiled target: /home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options profile 90 100 (PID 101436, part 1) Events recorded: Ir Events shown: Ir Event sort order: Ir Thresholds: 90 Include dirs: User annotated: Auto-annotation: on -------------------------------------------------------------------------------- Ir -------------------------------------------------------------------------------- 412,815,570 (100.0%) PROGRAM TOTALS -------------------------------------------------------------------------------- Ir file:function -------------------------------------------------------------------------------- 113,484,610 (27.49%) ???:calternal_markdown_view::source::lines::{closure#0} [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 58,447,185 (14.16%) ???:calternal_markdown_view::lines::extract_with_limits [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 29,856,206 ( 7.23%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_line [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 26,222,428 ( 6.35%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_block [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 22,170,813 ( 5.37%) ???:<pulldown_cmark::parse::OffsetIter as core::iter::traits::iterator::Iterator>::next [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 16,560,061 ( 4.01%) ???:<&mut calternal_markdown_view::bounded::Count as core::io::write::Write>::write_all [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 14,602,883 ( 3.54%) ???:<calternal_markdown_view::bounded::Budget>::new [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 13,026,906 ( 3.16%) ???:regex_automata::hybrid::search::find_fwd [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 11,912,352 ( 2.89%) ???:<regex::regex::string::Regex>::find_at [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 10,966,075 ( 2.66%) ???:<str>::trim_matches::<calternal_markdown_view::lines::title_trim> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 10,716,504 ( 2.60%) ???:<u8 as core::slice::cmp::SliceContains>::slice_contains [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 10,081,719 ( 2.44%) ???:<pulldown_cmark::firstpass::FirstPass>::parse_paragraph [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 6,512,343 ( 1.58%) ???:__memset_avx2_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 6,187,462 ( 1.50%) ???:<pulldown_cmark::tree::Tree<pulldown_cmark::parse::Item>>::append [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 5,909,813 ( 1.43%) ???:pulldown_cmark::parse::scan_containers [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 5,511,484 ( 1.34%) ???:__memcpy_avx_unaligned_erms [/nix/store/h4wfwic161kxrr74jlzla5lsm28hgary-glibc-2.44-25/lib/libc.so.6] 5,479,856 ( 1.33%) ???:core::ptr::drop_glue::<pulldown_cmark::Event> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] 5,344,011 ( 1.29%) ???:<calternal_markdown_view::bounded::Budget>::finish::<calternal_markdown_view::lines::Extraction> [/home/claude/Developer/Worktrees/calendar-805-r1/target/release/examples/parser_options] ```

kayg commented

2026-11-02 12:42:06 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/round2-tested-files.sha256

1c8d1dc1b3bc3dd14aee4c01d29c6532cdf8d23cf3cac9d0f89b05a14ecd7386  .forgejo/workflows/markdown-contract.yml
aee19e7a7347eb4294fe4d22ffd0df30e50ab1d7009b959a5a40b1bde953525a  Cargo.lock
bc6d1d84a4338daaa0c6a3a69dbf6e8c5443392db665bfe43dfffbbd01e01624  contracts/markdown-extraction/rows.json
9840d38eb4130867222d95b8e082219f8a2f79cc59c73ceac56cb23631fa95a0  crates/markdown-view/Cargo.toml
6dc8efb2bf8070225a2b239df93e716b82181d5221a6db805ded14e10bde4300  crates/markdown-view/src/bounded.rs
9a353425d24f74f68f16af4e047f5358d5966b2bda33dae26f4e3bf867ac148d  crates/markdown-view/src/lib.rs
65f7d80479dc99e21d3a37f9b3c7c6656862e54658d644fdbc34f6c507c1313a  crates/markdown-view/src/lines.rs
4998cc8e19d7a67e6fbd59ad4364198fc21a3ed3ec833b4d5322d172f401c9be  crates/markdown-view/src/source.rs
bff9f3de425e2b73150346432927ba8652901fe708062ed0e0a8b41ed08f503a  crates/markdown-view/tests/limits.rs
49a7e49d879b6c6038a17605c5f7ed1f3c13295784e441ed4fc29684b686cfa4  crates/markdown-view/tests/memory.rs
4016d313d01efcd692329c10e48157e692978e46d5cab24976ed587f50da8db7  packages/editor/src/markdown/extraction.ts
0afb4251c3f9312197f97c787885e02ec7d7f1f555a444fb778f537529dc0b73  packages/editor/test/markdown-rows.test.ts
9458bfc94d317a19295a41c1445376c038f7a94b3a57219fcd756381c0faed66  scripts/check-markdown-memory.py

ADR 0003 evidence migration: `docs/plans/0067-slice2/round2-tested-files.sha256` Source commit: [195168428eb02cced156c9247a6b226f4b68bce0](https://git.kayg.org/kayg/calternal/commit/195168428eb02cced156c9247a6b226f4b68bce0); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text 1c8d1dc1b3bc3dd14aee4c01d29c6532cdf8d23cf3cac9d0f89b05a14ecd7386 .forgejo/workflows/markdown-contract.yml aee19e7a7347eb4294fe4d22ffd0df30e50ab1d7009b959a5a40b1bde953525a Cargo.lock bc6d1d84a4338daaa0c6a3a69dbf6e8c5443392db665bfe43dfffbbd01e01624 contracts/markdown-extraction/rows.json 9840d38eb4130867222d95b8e082219f8a2f79cc59c73ceac56cb23631fa95a0 crates/markdown-view/Cargo.toml 6dc8efb2bf8070225a2b239df93e716b82181d5221a6db805ded14e10bde4300 crates/markdown-view/src/bounded.rs 9a353425d24f74f68f16af4e047f5358d5966b2bda33dae26f4e3bf867ac148d crates/markdown-view/src/lib.rs 65f7d80479dc99e21d3a37f9b3c7c6656862e54658d644fdbc34f6c507c1313a crates/markdown-view/src/lines.rs 4998cc8e19d7a67e6fbd59ad4364198fc21a3ed3ec833b4d5322d172f401c9be crates/markdown-view/src/source.rs bff9f3de425e2b73150346432927ba8652901fe708062ed0e0a8b41ed08f503a crates/markdown-view/tests/limits.rs 49a7e49d879b6c6038a17605c5f7ed1f3c13295784e441ed4fc29684b686cfa4 crates/markdown-view/tests/memory.rs 4016d313d01efcd692329c10e48157e692978e46d5cab24976ed587f50da8db7 packages/editor/src/markdown/extraction.ts 0afb4251c3f9312197f97c787885e02ec7d7f1f555a444fb778f537529dc0b73 packages/editor/test/markdown-rows.test.ts 9458bfc94d317a19295a41c1445376c038f7a94b3a57219fcd756381c0faed66 scripts/check-markdown-memory.py ```

kayg commented

2026-11-02 12:42:07 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/rules-contract-evidence.txt

Source commit: 2c6cb7ab60a4349a981f46d6d4004e5341f15786; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead
evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead
FORMAT_EXIT=0
    Finished `test` profile [unoptimized + debuginfo] target(s) in 0.09s
     Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-2eee79445a5aa94f)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/contract.rs (target/debug/deps/contract-366ef1b510a86bf3)

running 5 tests
18/18 untouched Composer configuration titles match authored fixtures
test existing_composer_configuration_matches_authored_titles ... ok
13/13 authored link vectors passed
test authored_link_contract ... ok
[H~2~O](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("H~2~O")), End(Link), End(Paragraph)]
[x^2^](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("x^2")), Text(Borrowed("^")), End(Link), End(Paragraph)]
[~~old~~](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Start(Strikethrough), Text(Borrowed("old")), End(Strikethrough), End(Link), End(Paragraph)]
test raw_configured_parser_label_evidence ... ok
test original_line_contract_through_candidate ... ok
test authored_source_rows ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

   Doc-tests calternal_markdown_view

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

NATIVE_EXIT=0
    Finished `test` profile [unoptimized + debuginfo] target(s) in 0.01s
     Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef)

running 1 test
70/70 pulldown line vectors passed
test tests::shared_line_vectors ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

   Doc-tests calendar_0067_parity_spike

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

SEED_EXIT=0
    Finished `test` profile [unoptimized + debuginfo] target(s) in 0.13s
     Running unittests src/lib.rs (target/debug/deps/calternal_tasks-1cfffd559dae52d1)

running 12 tests
composer "ordinary-source": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "composer-person-link": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
composer "unicode-byte-span": [240, 159, 152, 128, 32, 99, 97, 102, 195, 169]
composer "composer-formatting-outside-link": [42, 42, 77, 101, 101, 116, 42, 42, 32, 80, 114, 105, 121, 97, 32, 116, 111, 100, 97, 121]
test create_tests::concurrent_creates_name_clash_and_new_daily_undo ... ok
composer "composer-emphasis-in-label": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97]
test create_tests::create_ack_can_be_undone_immediately_and_reuses_daily_placeholder ... ok
composer "composer-code-in-label": [77, 101, 101, 116, 32, 97, 32, 98]
composer "code-does-not-create-link": [96, 91, 80, 114, 105, 121, 97, 93, 40, 60, 112, 46, 109, 100, 62, 41, 96]
composer "two-links": [65, 32, 97, 110, 100, 32, 66]
composer "emoji-source-label": [77, 101, 101, 116, 32, 58, 117, 110, 107, 110, 111, 119, 110, 95, 78, 65, 77, 69, 58]
composer "composer-subscript-compatibility": [77, 101, 101, 116, 32, 72, 126, 50, 126, 79]
composer "composer-superscript-compatibility": [77, 101, 101, 116, 32, 120, 94, 50, 94]
test create_tests::create_isolation_and_idempotent_retry ... ok
composer "composer-strikethrough-compatibility": [77, 101, 101, 116, 32, 126, 126, 111, 108, 100, 126, 126]
composer "literal-format-0xfeff": [239, 187, 191, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 239, 187, 191]
composer "literal-format-0x200b": [226, 128, 139, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 128, 139]
test create_tests::daily_uses_device_offset_and_invalid_slots_write_nothing ... ok
test create_tests::create_and_undo_restore_daily_bytes_and_remove_task ... ok
test create_tests::only_links_are_reduced_in_the_note_title ... ok
composer "literal-format-0x200c": [226, 128, 140, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 128, 140]
composer "literal-format-0x200d": [226, 128, 141, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 128, 141]
test create_tests::description_heading_keeps_the_given_task_title ... ok
composer "literal-format-0x2060": [226, 129, 160, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 129, 160]
test create_tests::composer_authored_title_corpus ... ok
test create_tests::opening_today_then_creating_tasks_scopes_keys_and_retries ... ok
test create_tests::undo_new_daily_capture_keeps_later_prose_and_removes_its_lines ... ok
test create_tests::repeated_create_undo_recreates_daily_with_retained_history ... ok
test create_tests::create_and_undo_one_mb_daily ... ok

test result: ok. 12 passed; 0 failed; 0 ignored; 0 measured; 30 filtered out; finished in 1.21s

COMPOSER_EXIT=0
bun install v1.4.2 (744846f84)

Checked 132 installs across 166 packages (no changes) [5.00ms]
INSTALL_EXIT=0
$ tsc --noEmit
TYPES_EXIT=0

[1m[30m[46m RUN [49m[39m[22m [36mv5.0.4 [39m[90m/home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/packages/editor[39m

 [32m✓[39m test/markdown-contract.test.ts [2m([22m[2m13 tests[22m[2m)[22m[32m 28[2mms[22m[39m
 [32m✓[39m test/calendar-0067.spike.test.ts [2m([22m[2m70 tests[22m[2m)[22m[32m 53[2mms[22m[39m
 [32m✓[39m test/markdown-rows.test.ts [2m([22m[2m116 tests[22m[2m)[22m[32m 82[2mms[22m[39m

[2m Test Files [22m [1m[32m3 passed[39m[22m[90m (3)[39m
[2m      Tests [22m [1m[32m199 passed[39m[22m[90m (199)[39m
[2m   Start at [22m 20:03:54
[2m   Duration [22m 1.09s[2m (environment 71%, import 13%, tests 7%, setup 5%, transform 2%, worker 2%)[22m

EDITOR_EXIT=0

[1m[30m[46m RUN [49m[39m[22m [36mv5.0.4 [39m[90m/home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/apps/web[39m

 [32m✓[39m src/composer/create.test.ts [2m([22m[2m5 tests[22m[2m)[22m[32m 28[2mms[22m[39m

[2m Test Files [22m [1m[32m1 passed[39m[22m[90m (1)[39m
[2m      Tests [22m [1m[32m5 passed[39m[22m[90m (5)[39m
[2m   Start at [22m 20:03:55
[2m   Duration [22m 286ms[2m (import 33%, transform 32%, tests 28%, worker 7%)[22m

COMPOSER_WEB_EXIT=0

ADR 0003 evidence migration: `docs/plans/0067-slice2/rules-contract-evidence.txt` Source commit: [2c6cb7ab60a4349a981f46d6d4004e5341f15786](https://git.kayg.org/kayg/calternal/commit/2c6cb7ab60a4349a981f46d6d4004e5341f15786); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text evaluation warning: stdenv.isLinux is deprecated, use stdenv.hostPlatform.isLinux instead evaluation warning: stdenv.isDarwin is deprecated, use stdenv.hostPlatform.isDarwin instead FORMAT_EXIT=0 Finished `test` profile [unoptimized + debuginfo] target(s) in 0.09s Running unittests src/lib.rs (target/debug/deps/calternal_markdown_view-2eee79445a5aa94f) running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s Running tests/contract.rs (target/debug/deps/contract-366ef1b510a86bf3) running 5 tests 18/18 untouched Composer configuration titles match authored fixtures test existing_composer_configuration_matches_authored_titles ... ok 13/13 authored link vectors passed test authored_link_contract ... ok [H~2~O](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("H~2~O")), End(Link), End(Paragraph)] [x^2^](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Text(Borrowed("x^2")), Text(Borrowed("^")), End(Link), End(Paragraph)] [~~old~~](<p.md>): [Start(Paragraph), Start(Link { link_type: Inline, dest_url: Borrowed("p.md"), title: Borrowed(""), id: Borrowed("") }), Start(Strikethrough), Text(Borrowed("old")), End(Strikethrough), End(Link), End(Paragraph)] test raw_configured_parser_label_evidence ... ok test original_line_contract_through_candidate ... ok test authored_source_rows ... ok test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Doc-tests calternal_markdown_view running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s NATIVE_EXIT=0 Finished `test` profile [unoptimized + debuginfo] target(s) in 0.01s Running unittests src/lib.rs (docs/plans/0067-vectors/rust/target/debug/deps/calendar_0067_parity_spike-b1d7c598b16993ef) running 1 test 70/70 pulldown line vectors passed test tests::shared_line_vectors ... ok test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s Doc-tests calendar_0067_parity_spike running 0 tests test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s SEED_EXIT=0 Finished `test` profile [unoptimized + debuginfo] target(s) in 0.13s Running unittests src/lib.rs (target/debug/deps/calternal_tasks-1cfffd559dae52d1) running 12 tests composer "ordinary-source": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "composer-person-link": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] composer "unicode-byte-span": [240, 159, 152, 128, 32, 99, 97, 102, 195, 169] composer "composer-formatting-outside-link": [42, 42, 77, 101, 101, 116, 42, 42, 32, 80, 114, 105, 121, 97, 32, 116, 111, 100, 97, 121] test create_tests::concurrent_creates_name_clash_and_new_daily_undo ... ok composer "composer-emphasis-in-label": [77, 101, 101, 116, 32, 80, 114, 105, 121, 97] test create_tests::create_ack_can_be_undone_immediately_and_reuses_daily_placeholder ... ok composer "composer-code-in-label": [77, 101, 101, 116, 32, 97, 32, 98] composer "code-does-not-create-link": [96, 91, 80, 114, 105, 121, 97, 93, 40, 60, 112, 46, 109, 100, 62, 41, 96] composer "two-links": [65, 32, 97, 110, 100, 32, 66] composer "emoji-source-label": [77, 101, 101, 116, 32, 58, 117, 110, 107, 110, 111, 119, 110, 95, 78, 65, 77, 69, 58] composer "composer-subscript-compatibility": [77, 101, 101, 116, 32, 72, 126, 50, 126, 79] composer "composer-superscript-compatibility": [77, 101, 101, 116, 32, 120, 94, 50, 94] test create_tests::create_isolation_and_idempotent_retry ... ok composer "composer-strikethrough-compatibility": [77, 101, 101, 116, 32, 126, 126, 111, 108, 100, 126, 126] composer "literal-format-0xfeff": [239, 187, 191, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 239, 187, 191] composer "literal-format-0x200b": [226, 128, 139, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 128, 139] test create_tests::daily_uses_device_offset_and_invalid_slots_write_nothing ... ok test create_tests::create_and_undo_restore_daily_bytes_and_remove_task ... ok test create_tests::only_links_are_reduced_in_the_note_title ... ok composer "literal-format-0x200c": [226, 128, 140, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 128, 140] composer "literal-format-0x200d": [226, 128, 141, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 128, 141] test create_tests::description_heading_keeps_the_given_task_title ... ok composer "literal-format-0x2060": [226, 129, 160, 77, 101, 101, 116, 32, 80, 114, 105, 121, 97, 226, 129, 160] test create_tests::composer_authored_title_corpus ... ok test create_tests::opening_today_then_creating_tasks_scopes_keys_and_retries ... ok test create_tests::undo_new_daily_capture_keeps_later_prose_and_removes_its_lines ... ok test create_tests::repeated_create_undo_recreates_daily_with_retained_history ... ok test create_tests::create_and_undo_one_mb_daily ... ok test result: ok. 12 passed; 0 failed; 0 ignored; 0 measured; 30 filtered out; finished in 1.21s COMPOSER_EXIT=0 bun install v1.4.2 (744846f84) Checked 132 installs across 166 packages (no changes) [5.00ms] INSTALL_EXIT=0 $ tsc --noEmit TYPES_EXIT=0 [1m[30m[46m RUN [49m[39m[22m [36mv5.0.4 [39m[90m/home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/packages/editor[39m [32m✓[39m test/markdown-contract.test.ts [2m([22m[2m13 tests[22m[2m)[22m[32m 28[2mms[22m[39m [32m✓[39m test/calendar-0067.spike.test.ts [2m([22m[2m70 tests[22m[2m)[22m[32m 53[2mms[22m[39m [32m✓[39m test/markdown-rows.test.ts [2m([22m[2m116 tests[22m[2m)[22m[32m 82[2mms[22m[39m [2m Test Files [22m [1m[32m3 passed[39m[22m[90m (3)[39m [2m Tests [22m [1m[32m199 passed[39m[22m[90m (199)[39m [2m Start at [22m 20:03:54 [2m Duration [22m 1.09s[2m (environment 71%, import 13%, tests 7%, setup 5%, transform 2%, worker 2%)[22m EDITOR_EXIT=0 [1m[30m[46m RUN [49m[39m[22m [36mv5.0.4 [39m[90m/home/claude/Developer/Worktrees/calendar-v2-s2-805-verify/apps/web[39m [32m✓[39m src/composer/create.test.ts [2m([22m[2m5 tests[22m[2m)[22m[32m 28[2mms[22m[39m [2m Test Files [22m [1m[32m1 passed[39m[22m[90m (1)[39m [2m Tests [22m [1m[32m5 passed[39m[22m[90m (5)[39m [2m Start at [22m 20:03:55 [2m Duration [22m 286ms[2m (import 33%, transform 32%, tests 28%, worker 7%)[22m COMPOSER_WEB_EXIT=0 ```

kayg commented

2026-11-02 12:42:07 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/rules-red-evidence.txt

Original file (32,927 bytes): rules-red-evidence.txt

ADR 0003 evidence migration: `docs/plans/0067-slice2/rules-red-evidence.txt` Source commit: [2c6cb7ab60a4349a981f46d6d4004e5341f15786](https://git.kayg.org/kayg/calternal/commit/2c6cb7ab60a4349a981f46d6d4004e5341f15786); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. Original file (32,927 bytes): [rules-red-evidence.txt](https://git.kayg.org/attachments/8b5e6362-d6d4-4656-8de5-774a64993d56)

kayg commented

2026-11-02 12:42:32 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/README.md

Original file (72,984 bytes): README.md

ADR 0003 evidence migration: `docs/plans/0067-slice2/README.md` Source commit: [195168428eb02cced156c9247a6b226f4b68bce0](https://git.kayg.org/kayg/calternal/commit/195168428eb02cced156c9247a6b226f4b68bce0); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. Original file (72,984 bytes): [README.md](https://git.kayg.org/attachments/9957684c-02b8-4219-93f4-ae3629593e66)

kayg commented

2026-11-02 12:43:55 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/baseline-worker.rs

//! Unconnected native worker candidate. Supervisor installs OS limits first.
//! Protocol: big-endian u32 byte length, then raw source / JSON outcome bytes.
use std::io::{self, Read, Write};
use calternal_markdown_view::extract;

fn main() -> io::Result<()> {
    let mut input = io::stdin().lock();
    let mut output = io::stdout().lock();
    output.write_all(b"READY")?;
    output.flush()?;
    loop {
        let mut header = [0; 4];
        match input.read_exact(&mut header) {
            Ok(()) => (),
            Err(e) if e.kind() == io::ErrorKind::UnexpectedEof => return Ok(()),
            Err(e) => return Err(e),
        }
        let size = u32::from_be_bytes(header) as usize;
        // Do not allocate an attacker-controlled frame. Supervisor replaces this
        // worker after protocol refusal, so the oversized body need not drain.
        if size >= 8 * 1024 * 1024 {
            send(&mut output, &serde_json::json!({"status":"unresolved", "reason":"source-limit"}))?;
            return Ok(());
        }
        let mut source = vec![0; size];
        input.read_exact(&mut source)?;
        match std::env::args().nth(1).as_deref() {
            Some("--fault-oom") => { std::hint::black_box(vec![1u8; 256 * 1024 * 1024]); }
            Some("--fault-stack") => { std::hint::black_box(stack(1024)); }
            _ => (),
        }
        let outcome = match std::str::from_utf8(&source).ok().and_then(|text| extract(text).ok()) {
            Some(value) => serde_json::json!({"status":"accepted", "value":value}),
            None => serde_json::json!({"status":"unresolved", "reason":"baseline-refusal"}),
        };
        send(&mut output, &outcome)?;
        // source, AST, extraction and encoded result all drop before next read.
    }
}
fn send(output: &mut impl Write, outcome: &serde_json::Value) -> io::Result<()> {
    let bytes = serde_json::to_vec(outcome)?;
    output.write_all(&(bytes.len() as u32).to_be_bytes())?;
    output.write_all(&bytes)?;
    output.flush()
}

#[inline(never)]
fn stack(n: usize) -> usize {
    let bytes = [1u8; 64 * 1024];
    let child = if n == 0 { 0 } else { stack(n - 1) };
    child + usize::from(std::hint::black_box(&bytes)[n % bytes.len()])
}

ADR 0003 evidence migration: `docs/plans/0067-slice2/baseline-worker.rs` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text //! Unconnected native worker candidate. Supervisor installs OS limits first. //! Protocol: big-endian u32 byte length, then raw source / JSON outcome bytes. use std::io::{self, Read, Write}; use calternal_markdown_view::extract; fn main() -> io::Result<()> { let mut input = io::stdin().lock(); let mut output = io::stdout().lock(); output.write_all(b"READY")?; output.flush()?; loop { let mut header = [0; 4]; match input.read_exact(&mut header) { Ok(()) => (), Err(e) if e.kind() == io::ErrorKind::UnexpectedEof => return Ok(()), Err(e) => return Err(e), } let size = u32::from_be_bytes(header) as usize; // Do not allocate an attacker-controlled frame. Supervisor replaces this // worker after protocol refusal, so the oversized body need not drain. if size >= 8 * 1024 * 1024 { send(&mut output, &serde_json::json!({"status":"unresolved", "reason":"source-limit"}))?; return Ok(()); } let mut source = vec![0; size]; input.read_exact(&mut source)?; match std::env::args().nth(1).as_deref() { Some("--fault-oom") => { std::hint::black_box(vec![1u8; 256 * 1024 * 1024]); } Some("--fault-stack") => { std::hint::black_box(stack(1024)); } _ => (), } let outcome = match std::str::from_utf8(&source).ok().and_then(|text| extract(text).ok()) { Some(value) => serde_json::json!({"status":"accepted", "value":value}), None => serde_json::json!({"status":"unresolved", "reason":"baseline-refusal"}), }; send(&mut output, &outcome)?; // source, AST, extraction and encoded result all drop before next read. } } fn send(output: &mut impl Write, outcome: &serde_json::Value) -> io::Result<()> { let bytes = serde_json::to_vec(outcome)?; output.write_all(&(bytes.len() as u32).to_be_bytes())?; output.write_all(&bytes)?; output.flush() } #[inline(never)] fn stack(n: usize) -> usize { let bytes = [1u8; 64 * 1024]; let child = if n == 0 { 0 } else { stack(n - 1) }; child + usize::from(std::hint::black_box(&bytes)[n % bytes.len()]) } ```

kayg commented

2026-11-02 12:43:56 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/check-native-limits.py

Source commit: 90b9e22ad7624549389362d884959ea9579c418b; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

#!/usr/bin/env python3
"""Native candidate gate, Linux only. No production worker/pool pass is implied.

Build the release example on a dev VM through slot. Run this on a locked perf VM
after a SHA-256 checked relay when using elapsed times as evidence. Every
probe starts under OS address-space/stack/CPU limits; timeout kills and reaps
only its own child. Report every assertion failure. No timing pass is claimed
on this shared build VM. The perf VM is reserved for subsequent scale gates.
"""
import json
import subprocess
import sys
import time

SOURCE = 8 * 1024 * 1024
DEADLINE = 0.050
BINARY = sys.argv[1]
failures = []


def probe(name, source, expected, *, deadline=DEADLINE, allowed_unresolved=None):
    started = time.monotonic()
    try:
        result = subprocess.run(
            ["prlimit", "--as=268435456", "--stack=8388608", "--cpu=1", "--", BINARY],
            input=source, stdout=subprocess.PIPE, stderr=subprocess.PIPE,
            timeout=deadline, check=False,
        )
    except subprocess.TimeoutExpired:
        failures.append(name)
        print(f"FAIL {name}: no adapter result before {deadline * 1000:g} ms; child killed/reaped", flush=True)
        return
    elapsed = (time.monotonic() - started) * 1000
    if result.returncode:
        failures.append(name)
        print(f"FAIL {name}: worker exit={result.returncode}, stderr={result.stderr[-1024:]!r}", flush=True)
        return
    if len(result.stdout) > 16 * 1024 * 1024:
        failures.append(name)
        print(f"FAIL {name}: encoded result exceeds 16 MiB", flush=True)
        return
    value = json.loads(result.stdout)
    status = (value.get("status"), value.get("reason"))
    if status != expected and not (allowed_unresolved and status == ("unresolved", allowed_unresolved)):
        rows = value.get("value", {}).get("rows", [])
        failures.append(name)
        print(f"FAIL {name}: expected {expected}, got {status}, rows={len(rows)}, bytes={len(source)}, elapsed_ms={elapsed:.3f}", flush=True)
        return
    print(f"PASS {name}: {status}, bytes={len(source)}, elapsed_ms={elapsed:.3f}", flush=True)


probe("ordinary", b"- 09:00 Meet **#area/work**", ("accepted", None))
for name, invalid in (("lead", b"\xff"), ("continuation", b"\x80"), ("truncated", b"\xe2\x82")):
    probe(f"invalid-utf8-{name}", b"- 09:00 " + invalid, ("unresolved", "invalid-utf8"))
# D counts blockquotes plus list-item nesting, not List + Item AST wrappers.
# The final task adds one list level to each quote-only boundary source.
for depth in (255, 256, 257):
    expected = ("accepted", None) if depth < 256 else ("unresolved", "depth-limit")
    probe(f"depth-{depth}", b"> " * (depth - 1) + b"- [ ] todo", expected)
probe("original-depth-repro", b"> " * 257 + b"- [ ] todo", ("unresolved", "depth-limit"))
# Seeded mixed single-line containers, plus multiline list continuations. These
# sources construct known depths independently of the parser under test.
import random
rng = random.Random(805)
for case in range(24):
    depth = rng.choice((255, 256, 257))
    prefix = b"".join(rng.choice((b"> ", b"- ", b"1. ", b"+ ")) for _ in range(depth - 1))
    actual = prefix.count(b">") + prefix.count(b"- ") + prefix.count(b"1. ") + prefix.count(b"+ ") + 1
    expected = ("accepted", None) if actual < 256 else ("unresolved", "depth-limit")
    probe(f"mixed-{case}-depth-{actual}", prefix + b"- [ ] todo", expected)
for depth in (255, 256, 257):
    source = b"".join(b"  " * i + b"- item\n" for i in range(depth - 1))
    source += b"  " * (depth - 1) + b"- [ ] todo"
    expected = ("accepted", None) if depth < 256 else ("unresolved", "depth-limit")
    probe(f"list-depth-{depth}", source, expected)
# Reached only if all preceding gates pass. Extra allowance here is for transfer
# and launch; a real persistent supervisor still has to enforce its 50 ms parse
# deadline. This one-shot probe cannot certify that supervisor contract.
probe("huge-below-limit", b"x" * (SOURCE - 1), ("accepted", None), deadline=1)
probe("huge-at-limit", b"x" * SOURCE, ("unresolved", "source-limit"), deadline=1)
probe("huge-over-limit", b"x" * (SOURCE + 1), ("unresolved", "source-limit"), deadline=1)

# 147,464 bytes, one event row, ~65k parser nodes: below the byte/node/row/
# result caps. Many ordinary tag wrappers exercise extraction work, not just
# launch/transfer or parser construction. A supervisor may refuse on deadline;
# killing/reaping a hung one-shot child alone is not a defined adapter outcome.
probe("formatted-tag-storm-deadline", b"- 09:00 " + b"**#tag** " * 16_384,
      ("accepted", None), allowed_unresolved="timeout")

# One paragraph, one initial text leaf, then distinct inline-HTML leaves.
# Prefix prose prevents HTML-block parsing. The two inputs contain exactly
# 250,000 / 250,001 semantic nodes (End events are not nodes). Excluded HTML is
# still parser work and must count towards B. This is a cap/outcome smoke test,
# with launch/transfer allowance, not the persistent 50 ms deadline proof.
probe("nodes-below-limit", b"x " + b"<i>" * 249_997,
      ("accepted", None), deadline=1)
probe("nodes-at-limit", b"x " + b"<i>" * 249_998,
      ("unresolved", "node-limit"), deadline=1)
probe("nodes-over-limit", b"x " + b"<i>" * 249_999,
      ("unresolved", "node-limit"), deadline=1)

print(f"SUMMARY failures={len(failures)} names={failures}", flush=True)
sys.exit(bool(failures))

ADR 0003 evidence migration: `docs/plans/0067-slice2/check-native-limits.py` Source commit: [90b9e22ad7624549389362d884959ea9579c418b](https://git.kayg.org/kayg/calternal/commit/90b9e22ad7624549389362d884959ea9579c418b); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text #!/usr/bin/env python3 """Native candidate gate, Linux only. No production worker/pool pass is implied. Build the release example on a dev VM through slot. Run this on a locked perf VM after a SHA-256 checked relay when using elapsed times as evidence. Every probe starts under OS address-space/stack/CPU limits; timeout kills and reaps only its own child. Report every assertion failure. No timing pass is claimed on this shared build VM. The perf VM is reserved for subsequent scale gates. """ import json import subprocess import sys import time SOURCE = 8 * 1024 * 1024 DEADLINE = 0.050 BINARY = sys.argv[1] failures = [] def probe(name, source, expected, *, deadline=DEADLINE, allowed_unresolved=None): started = time.monotonic() try: result = subprocess.run( ["prlimit", "--as=268435456", "--stack=8388608", "--cpu=1", "--", BINARY], input=source, stdout=subprocess.PIPE, stderr=subprocess.PIPE, timeout=deadline, check=False, ) except subprocess.TimeoutExpired: failures.append(name) print(f"FAIL {name}: no adapter result before {deadline * 1000:g} ms; child killed/reaped", flush=True) return elapsed = (time.monotonic() - started) * 1000 if result.returncode: failures.append(name) print(f"FAIL {name}: worker exit={result.returncode}, stderr={result.stderr[-1024:]!r}", flush=True) return if len(result.stdout) > 16 * 1024 * 1024: failures.append(name) print(f"FAIL {name}: encoded result exceeds 16 MiB", flush=True) return value = json.loads(result.stdout) status = (value.get("status"), value.get("reason")) if status != expected and not (allowed_unresolved and status == ("unresolved", allowed_unresolved)): rows = value.get("value", {}).get("rows", []) failures.append(name) print(f"FAIL {name}: expected {expected}, got {status}, rows={len(rows)}, bytes={len(source)}, elapsed_ms={elapsed:.3f}", flush=True) return print(f"PASS {name}: {status}, bytes={len(source)}, elapsed_ms={elapsed:.3f}", flush=True) probe("ordinary", b"- 09:00 Meet **#area/work**", ("accepted", None)) for name, invalid in (("lead", b"\xff"), ("continuation", b"\x80"), ("truncated", b"\xe2\x82")): probe(f"invalid-utf8-{name}", b"- 09:00 " + invalid, ("unresolved", "invalid-utf8")) # D counts blockquotes plus list-item nesting, not List + Item AST wrappers. # The final task adds one list level to each quote-only boundary source. for depth in (255, 256, 257): expected = ("accepted", None) if depth < 256 else ("unresolved", "depth-limit") probe(f"depth-{depth}", b"> " * (depth - 1) + b"- [ ] todo", expected) probe("original-depth-repro", b"> " * 257 + b"- [ ] todo", ("unresolved", "depth-limit")) # Seeded mixed single-line containers, plus multiline list continuations. These # sources construct known depths independently of the parser under test. import random rng = random.Random(805) for case in range(24): depth = rng.choice((255, 256, 257)) prefix = b"".join(rng.choice((b"> ", b"- ", b"1. ", b"+ ")) for _ in range(depth - 1)) actual = prefix.count(b">") + prefix.count(b"- ") + prefix.count(b"1. ") + prefix.count(b"+ ") + 1 expected = ("accepted", None) if actual < 256 else ("unresolved", "depth-limit") probe(f"mixed-{case}-depth-{actual}", prefix + b"- [ ] todo", expected) for depth in (255, 256, 257): source = b"".join(b" " * i + b"- item\n" for i in range(depth - 1)) source += b" " * (depth - 1) + b"- [ ] todo" expected = ("accepted", None) if depth < 256 else ("unresolved", "depth-limit") probe(f"list-depth-{depth}", source, expected) # Reached only if all preceding gates pass. Extra allowance here is for transfer # and launch; a real persistent supervisor still has to enforce its 50 ms parse # deadline. This one-shot probe cannot certify that supervisor contract. probe("huge-below-limit", b"x" * (SOURCE - 1), ("accepted", None), deadline=1) probe("huge-at-limit", b"x" * SOURCE, ("unresolved", "source-limit"), deadline=1) probe("huge-over-limit", b"x" * (SOURCE + 1), ("unresolved", "source-limit"), deadline=1) # 147,464 bytes, one event row, ~65k parser nodes: below the byte/node/row/ # result caps. Many ordinary tag wrappers exercise extraction work, not just # launch/transfer or parser construction. A supervisor may refuse on deadline; # killing/reaping a hung one-shot child alone is not a defined adapter outcome. probe("formatted-tag-storm-deadline", b"- 09:00 " + b"**#tag** " * 16_384, ("accepted", None), allowed_unresolved="timeout") # One paragraph, one initial text leaf, then distinct inline-HTML leaves. # Prefix prose prevents HTML-block parsing. The two inputs contain exactly # 250,000 / 250,001 semantic nodes (End events are not nodes). Excluded HTML is # still parser work and must count towards B. This is a cap/outcome smoke test, # with launch/transfer allowance, not the persistent 50 ms deadline proof. probe("nodes-below-limit", b"x " + b"<i>" * 249_997, ("accepted", None), deadline=1) probe("nodes-at-limit", b"x " + b"<i>" * 249_998, ("unresolved", "node-limit"), deadline=1) probe("nodes-over-limit", b"x " + b"<i>" * 249_999, ("unresolved", "node-limit"), deadline=1) print(f"SUMMARY failures={len(failures)} names={failures}", flush=True) sys.exit(bool(failures)) ```

kayg commented

2026-11-02 12:43:56 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/check-worker-package.py

Source commit: 73a24063d9f7c318e3fc3f75060ab4cb0745aded; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

#!/usr/bin/env python3
"""Functional packaging only. No latency claim on a build VM."""
import importlib.util
import json
import platform
import sys

spec = importlib.util.spec_from_file_location("pool", sys.argv[1])
module = importlib.util.module_from_spec(spec)
spec.loader.exec_module(module)
binary = sys.argv[2]
results = {"os": platform.system(), "arch": platform.machine()}
for fault in [None, "oom", "stack"]:
    worker = module.Worker(binary, deadline=1, fault=fault)
    try:
        result = worker.parse(b"- [ ] hello\n- 09:00 event")
        if fault:
            assert result == module.refused("worker-death"), result
            assert worker.parse(b"- [ ] recovered")["status"] == "accepted"
        else:
            assert result["status"] == "accepted" and len(result["value"]["rows"]) == 2, result
            assert worker.parse(b"\xff") == module.refused("invalid-utf8")
        results[fault or "ordinary"] = "pass"
    finally:
        worker.close()
print(json.dumps(results), flush=True)

ADR 0003 evidence migration: `docs/plans/0067-slice2/check-worker-package.py` Source commit: [73a24063d9f7c318e3fc3f75060ab4cb0745aded](https://git.kayg.org/kayg/calternal/commit/73a24063d9f7c318e3fc3f75060ab4cb0745aded); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text #!/usr/bin/env python3 """Functional packaging only. No latency claim on a build VM.""" import importlib.util import json import platform import sys spec = importlib.util.spec_from_file_location("pool", sys.argv[1]) module = importlib.util.module_from_spec(spec) spec.loader.exec_module(module) binary = sys.argv[2] results = {"os": platform.system(), "arch": platform.machine()} for fault in [None, "oom", "stack"]: worker = module.Worker(binary, deadline=1, fault=fault) try: result = worker.parse(b"- [ ] hello\n- 09:00 event") if fault: assert result == module.refused("worker-death"), result assert worker.parse(b"- [ ] recovered")["status"] == "accepted" else: assert result["status"] == "accepted" and len(result["value"]["rows"]) == 2, result assert worker.parse(b"\xff") == module.refused("invalid-utf8") results[fault or "ordinary"] = "pass" finally: worker.close() print(json.dumps(results), flush=True) ```

kayg commented

2026-11-02 12:43:58 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/compare-optimized.py

Source commit: 343acaa60bab1995b7ed450865c506a9bbdfdc0e; snapshot from dev 55f35fcd. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable.

#!/usr/bin/env python3
"""Run only under perf-lock; hash-verified dev-VM release artifacts required."""
import ast
import hashlib
import json
import os
from pathlib import Path
import resource
import subprocess
import sys
import tempfile

root = Path(sys.argv[1]).resolve()

# Use the original independently declared fixture without executing its gates.
tree = ast.parse((root / "gate-pool.py").read_text())
fixture_def = next(n for n in tree.body if isinstance(n, ast.FunctionDef) and n.name == "fixture")
scope = {}
exec(compile(ast.Module(body=[fixture_def], type_ignores=[]), "gate-pool.py", "exec"), scope)
fixture = scope["fixture"]
for index in range(100):
    actual = subprocess.check_output([root / "options-head", "fixture", str(index)])
    assert actual == fixture(index)[0], index
hashes = []
for shard in range(2):
    digest = hashlib.sha256()
    for index in range(shard, 100_000, 2):
        digest.update(fixture(index)[0])
    hashes.append(digest.hexdigest())
print("FIXTURE", json.dumps({"byte_identity_100_variants": True, "shard_sha256": hashes}), flush=True)


def run(binary, mode, count):
    with tempfile.TemporaryFile() as output:
        child = subprocess.Popen([str(root / binary), mode, str(count),
                                  str(root / "markdown_pool.py"), str(root / "worker-head")], stdout=output)
        _, status, usage = os.wait4(child.pid, 0)
        child.returncode = os.waitstatus_to_exitcode(status)
        output.seek(0)
        text = output.read().decode()
    print("RAW", binary, text.strip(), flush=True)
    assert child.returncode == 0, child.returncode
    report = json.loads(text)
    assert report["notes"] == count and report["rows"] == count * 2
    assert report["edges"] == count // 2 and report["unresolved"] == 0
    assert report["bytes"] == 925_696_000 * count // 100_000
    report.update(cpu_s=usage.ru_utime + usage.ru_stime, maxrss_kib=usage.ru_maxrss)
    print("MEASURE", binary, json.dumps(report), flush=True)
    return report


reports = {}
for count in (10_000, 100_000):
    for name, binary, mode in (("baseline-direct", "options-baseline", "baseline"),
                               ("A", "options-head", "A")):
        reports[name, count] = run(binary, mode, count)
for name in ("A",):
    result = reports[name, 100_000]
    baseline = reports["baseline-direct", 100_000]
    buckets = []
    for before, after in zip(baseline["buckets"], result["buckets"]):
        added = after["p95_ms"] - before["p95_ms"]
        growth = after["p95_ms"] / before["p95_ms"] - 1
        buckets.append({"bytes": after["bytes"], "added_p95_ms": added, "growth": growth,
                        "added_pass": added <= 0.2, "growth_pass": growth <= 0.1})
    print("BUDGET", json.dumps({"option": name, "buckets": buckets,
          "vs_historical_12.614s": result["wall_s"] / 12.614 - 1,
          "vs_paired_direct": result["wall_s"] / baseline["wall_s"] - 1,
          "historical_wall_pass": result["wall_s"] <= 12.614 * 1.1,
          "paired_wall_pass": result["wall_s"] <= baseline["wall_s"] * 1.1}), flush=True)

    assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets
    assert result["wall_s"] <= baseline["wall_s"] * 1.1, "paired 100k wall budget"


def limits():
    resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_CORE, (0, 0))


for name, source in (("ordinary", b"- [ ] good\n- 09:00 event"),
                     ("depth255", b"> " * 254 + b"- [ ] good"),
                     ("depth256", b"> " * 255 + b"- [ ] good"),
                     ("long-paragraph", b"x" * (8 * 1024 * 1024 - 1)),
                     ("short-line-storm", b"\n" * (8 * 1024 * 1024 - 1)),
                     ("inline-storm", b"*x* " * (256 * 1024)),
                     ("source-limit", b"x" * (8 * 1024 * 1024))):
    try:
        child = subprocess.run([root / "in_process_containment"], input=source,
                               capture_output=True, timeout=3, preexec_fn=limits, check=False)
        print("CONTAINMENT", json.dumps({"probe": name, "bytes": len(source), "exit": child.returncode,
              "stdout": child.stdout.decode(), "stderr_tail": child.stderr[-512:].decode(errors="replace")}), flush=True)
    except subprocess.TimeoutExpired:
        print("CONTAINMENT", json.dumps({"probe": name, "external_kill_after_s": 3}), flush=True)

ADR 0003 evidence migration: `docs/plans/0067-slice2/compare-optimized.py` Source commit: [343acaa60bab1995b7ed450865c506a9bbdfdc0e](https://git.kayg.org/kayg/calternal/commit/343acaa60bab1995b7ed450865c506a9bbdfdc0e); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text #!/usr/bin/env python3 """Run only under perf-lock; hash-verified dev-VM release artifacts required.""" import ast import hashlib import json import os from pathlib import Path import resource import subprocess import sys import tempfile root = Path(sys.argv[1]).resolve() # Use the original independently declared fixture without executing its gates. tree = ast.parse((root / "gate-pool.py").read_text()) fixture_def = next(n for n in tree.body if isinstance(n, ast.FunctionDef) and n.name == "fixture") scope = {} exec(compile(ast.Module(body=[fixture_def], type_ignores=[]), "gate-pool.py", "exec"), scope) fixture = scope["fixture"] for index in range(100): actual = subprocess.check_output([root / "options-head", "fixture", str(index)]) assert actual == fixture(index)[0], index hashes = [] for shard in range(2): digest = hashlib.sha256() for index in range(shard, 100_000, 2): digest.update(fixture(index)[0]) hashes.append(digest.hexdigest()) print("FIXTURE", json.dumps({"byte_identity_100_variants": True, "shard_sha256": hashes}), flush=True) def run(binary, mode, count): with tempfile.TemporaryFile() as output: child = subprocess.Popen([str(root / binary), mode, str(count), str(root / "markdown_pool.py"), str(root / "worker-head")], stdout=output) _, status, usage = os.wait4(child.pid, 0) child.returncode = os.waitstatus_to_exitcode(status) output.seek(0) text = output.read().decode() print("RAW", binary, text.strip(), flush=True) assert child.returncode == 0, child.returncode report = json.loads(text) assert report["notes"] == count and report["rows"] == count * 2 assert report["edges"] == count // 2 and report["unresolved"] == 0 assert report["bytes"] == 925_696_000 * count // 100_000 report.update(cpu_s=usage.ru_utime + usage.ru_stime, maxrss_kib=usage.ru_maxrss) print("MEASURE", binary, json.dumps(report), flush=True) return report reports = {} for count in (10_000, 100_000): for name, binary, mode in (("baseline-direct", "options-baseline", "baseline"), ("A", "options-head", "A")): reports[name, count] = run(binary, mode, count) for name in ("A",): result = reports[name, 100_000] baseline = reports["baseline-direct", 100_000] buckets = [] for before, after in zip(baseline["buckets"], result["buckets"]): added = after["p95_ms"] - before["p95_ms"] growth = after["p95_ms"] / before["p95_ms"] - 1 buckets.append({"bytes": after["bytes"], "added_p95_ms": added, "growth": growth, "added_pass": added <= 0.2, "growth_pass": growth <= 0.1}) print("BUDGET", json.dumps({"option": name, "buckets": buckets, "vs_historical_12.614s": result["wall_s"] / 12.614 - 1, "vs_paired_direct": result["wall_s"] / baseline["wall_s"] - 1, "historical_wall_pass": result["wall_s"] <= 12.614 * 1.1, "paired_wall_pass": result["wall_s"] <= baseline["wall_s"] * 1.1}), flush=True) assert all(b["added_pass"] and b["growth_pass"] for b in buckets), buckets assert result["wall_s"] <= baseline["wall_s"] * 1.1, "paired 100k wall budget" def limits(): resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_CORE, (0, 0)) for name, source in (("ordinary", b"- [ ] good\n- 09:00 event"), ("depth255", b"> " * 254 + b"- [ ] good"), ("depth256", b"> " * 255 + b"- [ ] good"), ("long-paragraph", b"x" * (8 * 1024 * 1024 - 1)), ("short-line-storm", b"\n" * (8 * 1024 * 1024 - 1)), ("inline-storm", b"*x* " * (256 * 1024)), ("source-limit", b"x" * (8 * 1024 * 1024))): try: child = subprocess.run([root / "in_process_containment"], input=source, capture_output=True, timeout=3, preexec_fn=limits, check=False) print("CONTAINMENT", json.dumps({"probe": name, "bytes": len(source), "exit": child.returncode, "stdout": child.stdout.decode(), "stderr_tail": child.stderr[-512:].decode(errors="replace")}), flush=True) except subprocess.TimeoutExpired: print("CONTAINMENT", json.dumps({"probe": name, "external_kill_after_s": 3}), flush=True) ```

kayg commented

2026-11-02 12:43:59 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/compare-options.py

#!/usr/bin/env python3
"""Run only under perf-lock; hash-verified dev-VM release artifacts required."""
import ast
import hashlib
import json
import os
from pathlib import Path
import resource
import subprocess
import sys
import tempfile

root = Path(sys.argv[1]).resolve()

# Use the original independently declared fixture without executing its gates.
tree = ast.parse((root / "gate-pool.py").read_text())
fixture_def = next(n for n in tree.body if isinstance(n, ast.FunctionDef) and n.name == "fixture")
scope = {}
exec(compile(ast.Module(body=[fixture_def], type_ignores=[]), "gate-pool.py", "exec"), scope)
fixture = scope["fixture"]
for index in range(100):
    actual = subprocess.check_output([root / "options-head", "fixture", str(index)])
    assert actual == fixture(index)[0], index
hashes = []
for shard in range(2):
    digest = hashlib.sha256()
    for index in range(shard, 100_000, 2):
        digest.update(fixture(index)[0])
    hashes.append(digest.hexdigest())
print("FIXTURE", json.dumps({"byte_identity_100_variants": True, "shard_sha256": hashes}), flush=True)


def run(binary, mode, count):
    with tempfile.TemporaryFile() as output:
        child = subprocess.Popen([str(root / binary), mode, str(count),
                                  str(root / "markdown_pool.py"), str(root / "worker-head")], stdout=output)
        _, status, usage = os.wait4(child.pid, 0)
        child.returncode = os.waitstatus_to_exitcode(status)
        output.seek(0)
        text = output.read().decode()
    print("RAW", binary, text.strip(), flush=True)
    assert child.returncode == 0, child.returncode
    report = json.loads(text)
    assert report["notes"] == count and report["rows"] == count * 2
    assert report["edges"] == count // 2 and report["unresolved"] == 0
    assert report["bytes"] == 925_696_000 * count // 100_000
    report.update(cpu_s=usage.ru_utime + usage.ru_stime, maxrss_kib=usage.ru_maxrss)
    print("MEASURE", binary, json.dumps(report), flush=True)
    return report


reports = {}
for count in (10_000, 100_000):
    for name, binary, mode in (("baseline-direct", "options-baseline", "baseline"),
                               ("A", "options-head", "A"), ("C", "options-head", "C")):
        reports[name, count] = run(binary, mode, count)
for name in ("A", "C"):
    result = reports[name, 100_000]
    baseline = reports["baseline-direct", 100_000]
    buckets = []
    for before, after in zip(baseline["buckets"], result["buckets"]):
        added = after["p95_ms"] - before["p95_ms"]
        growth = after["p95_ms"] / before["p95_ms"] - 1
        buckets.append({"bytes": after["bytes"], "added_p95_ms": added, "growth": growth,
                        "pass": added <= 0.2 and growth <= 0.1})
    print("BUDGET", json.dumps({"option": name, "buckets": buckets,
          "vs_historical_12.614s": result["wall_s"] / 12.614 - 1,
          "vs_paired_direct": result["wall_s"] / baseline["wall_s"] - 1,
          "historical_wall_pass": result["wall_s"] <= 12.614 * 1.1,
          "paired_wall_pass": result["wall_s"] <= baseline["wall_s"] * 1.1}), flush=True)


def limits():
    resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_CORE, (0, 0))


for name, source in (("ordinary", b"- [ ] good\n- 09:00 event"),
                     ("depth255", b"> " * 254 + b"- [ ] good"),
                     ("depth256", b"> " * 255 + b"- [ ] good"),
                     ("long-paragraph", b"x" * (8 * 1024 * 1024 - 1)),
                     ("short-line-storm", b"\n" * (8 * 1024 * 1024 - 1)),
                     ("inline-storm", b"*x* " * (256 * 1024)),
                     ("source-limit", b"x" * (8 * 1024 * 1024))):
    try:
        child = subprocess.run([root / "in_process_containment"], input=source,
                               capture_output=True, timeout=3, preexec_fn=limits, check=False)
        print("CONTAINMENT", json.dumps({"probe": name, "bytes": len(source), "exit": child.returncode,
              "stdout": child.stdout.decode(), "stderr_tail": child.stderr[-512:].decode(errors="replace")}), flush=True)
    except subprocess.TimeoutExpired:
        print("CONTAINMENT", json.dumps({"probe": name, "external_kill_after_s": 3}), flush=True)

ADR 0003 evidence migration: `docs/plans/0067-slice2/compare-options.py` Source commit: [9064e29ac5ca9744cd369c1137a16bfd764e37d3](https://git.kayg.org/kayg/calternal/commit/9064e29ac5ca9744cd369c1137a16bfd764e37d3); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text #!/usr/bin/env python3 """Run only under perf-lock; hash-verified dev-VM release artifacts required.""" import ast import hashlib import json import os from pathlib import Path import resource import subprocess import sys import tempfile root = Path(sys.argv[1]).resolve() # Use the original independently declared fixture without executing its gates. tree = ast.parse((root / "gate-pool.py").read_text()) fixture_def = next(n for n in tree.body if isinstance(n, ast.FunctionDef) and n.name == "fixture") scope = {} exec(compile(ast.Module(body=[fixture_def], type_ignores=[]), "gate-pool.py", "exec"), scope) fixture = scope["fixture"] for index in range(100): actual = subprocess.check_output([root / "options-head", "fixture", str(index)]) assert actual == fixture(index)[0], index hashes = [] for shard in range(2): digest = hashlib.sha256() for index in range(shard, 100_000, 2): digest.update(fixture(index)[0]) hashes.append(digest.hexdigest()) print("FIXTURE", json.dumps({"byte_identity_100_variants": True, "shard_sha256": hashes}), flush=True) def run(binary, mode, count): with tempfile.TemporaryFile() as output: child = subprocess.Popen([str(root / binary), mode, str(count), str(root / "markdown_pool.py"), str(root / "worker-head")], stdout=output) _, status, usage = os.wait4(child.pid, 0) child.returncode = os.waitstatus_to_exitcode(status) output.seek(0) text = output.read().decode() print("RAW", binary, text.strip(), flush=True) assert child.returncode == 0, child.returncode report = json.loads(text) assert report["notes"] == count and report["rows"] == count * 2 assert report["edges"] == count // 2 and report["unresolved"] == 0 assert report["bytes"] == 925_696_000 * count // 100_000 report.update(cpu_s=usage.ru_utime + usage.ru_stime, maxrss_kib=usage.ru_maxrss) print("MEASURE", binary, json.dumps(report), flush=True) return report reports = {} for count in (10_000, 100_000): for name, binary, mode in (("baseline-direct", "options-baseline", "baseline"), ("A", "options-head", "A"), ("C", "options-head", "C")): reports[name, count] = run(binary, mode, count) for name in ("A", "C"): result = reports[name, 100_000] baseline = reports["baseline-direct", 100_000] buckets = [] for before, after in zip(baseline["buckets"], result["buckets"]): added = after["p95_ms"] - before["p95_ms"] growth = after["p95_ms"] / before["p95_ms"] - 1 buckets.append({"bytes": after["bytes"], "added_p95_ms": added, "growth": growth, "pass": added <= 0.2 and growth <= 0.1}) print("BUDGET", json.dumps({"option": name, "buckets": buckets, "vs_historical_12.614s": result["wall_s"] / 12.614 - 1, "vs_paired_direct": result["wall_s"] / baseline["wall_s"] - 1, "historical_wall_pass": result["wall_s"] <= 12.614 * 1.1, "paired_wall_pass": result["wall_s"] <= baseline["wall_s"] * 1.1}), flush=True) def limits(): resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_CORE, (0, 0)) for name, source in (("ordinary", b"- [ ] good\n- 09:00 event"), ("depth255", b"> " * 254 + b"- [ ] good"), ("depth256", b"> " * 255 + b"- [ ] good"), ("long-paragraph", b"x" * (8 * 1024 * 1024 - 1)), ("short-line-storm", b"\n" * (8 * 1024 * 1024 - 1)), ("inline-storm", b"*x* " * (256 * 1024)), ("source-limit", b"x" * (8 * 1024 * 1024))): try: child = subprocess.run([root / "in_process_containment"], input=source, capture_output=True, timeout=3, preexec_fn=limits, check=False) print("CONTAINMENT", json.dumps({"probe": name, "bytes": len(source), "exit": child.returncode, "stdout": child.stdout.decode(), "stderr_tail": child.stderr[-512:].decode(errors="replace")}), flush=True) except subprocess.TimeoutExpired: print("CONTAINMENT", json.dumps({"probe": name, "external_kill_after_s": 3}), flush=True) ```

kayg commented

2026-11-02 12:43:59 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/gate-pool.py

#!/usr/bin/env python3
"""Run all containment probes, then paired normal-corpus scale measurements.

Run only on a locked perf VM with hash-verified release workers. No fail-fast.
"""
import concurrent.futures
import hashlib
import importlib.util
import json
import os
from pathlib import Path
import resource
import signal
import sqlite3
import statistics
import sys
import tempfile
import time

spec = importlib.util.spec_from_file_location("markdown_pool", sys.argv[1])
pool_module = importlib.util.module_from_spec(spec)
spec.loader.exec_module(pool_module)
Worker = pool_module.Worker
pools = []
class Pool(pool_module.Pool):
    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        pools.append(self)
HEAD, BASE = sys.argv[2:4]
failures = []


def gate(name, run):
    try:
        result = run()
        print("PASS", name, json.dumps(result), flush=True)
    except Exception as error:
        failures.append(name)
        print("FAIL", name, repr(error), flush=True)
    finally:
        for pool in pools:
            pool.close()


def digest(source):
    return hashlib.sha256(source).hexdigest()


def containment():
    with tempfile.TemporaryDirectory() as directory:
        path, ledger = Path(directory) / "note.md", Path(directory) / "ledger.db"
        started = time.monotonic()
        pool = Pool(HEAD, ledger)
        startup = (time.monotonic() - started) * 1000
        assert startup <= 500, startup
        good = b"- [ ] keep me\n- 09:00 event"
        path.write_bytes(good)
        revision = digest(good)
        def submit(source):
            path.write_bytes(source)
            return pool.submit("one", "note", digest(source), path, True).result(5)
        result = submit(good)
        assert result["status"] == "accepted", result
        assert len(result["value"]["rows"]) == 2, result
        original = pool.db.execute("SELECT good FROM notes").fetchone()[0]
        probes = {}
        for name, source in [("invalid-utf8", b"\xff"), ("source-limit", b"x" * (8 * 1024 * 1024)),
                             ("depth-limit", b"> " * 256 + b"x")]:
            result = submit(source)
            assert result["status"] == "unresolved" and result["reason"] == name, result
            assert pool.db.execute("SELECT good FROM notes").fetchone()[0] == original
            assert path.read_bytes() == source
            probes[name] = result["reason"]
        for sig, expected in [(signal.SIGKILL, "worker-death"), (signal.SIGSTOP, "timeout")]:
            for worker in pool.workers:
                os.kill(worker.child.pid, sig)
            result = submit(good + str(sig).encode())
            assert result["status"] == "unresolved" and result["reason"] == expected, result
            assert pool.db.execute("SELECT good FROM notes").fetchone()[0] == original
            probes[expected] = result["reason"]
            # Replace the idle killed/stopped worker too, before proving recovery.
            for worker in pool.workers:
                worker.replace()
            for worker in pool.workers:
                assert worker.parse(good)["status"] == "accepted"
        for fault in ["oom", "stack"]:
            worker = Worker(HEAD, fault=fault)
            try:
                result = worker.parse(good)
                assert result == pool_module.refused("worker-death"), result
                assert worker.parse(good)["status"] == "accepted"
                probes[fault] = "worker-death/replaced"
            finally:
                worker.close()
        # Hold dispatch under the same lock to deterministically fill/coalesce.
        path.write_bytes(good)
        with pool.condition:
            futures = [pool.submit(str(i % 3), str(i), revision, path) for i in range(56)]
            assert pool.submit("other", "over-background", revision, path).result()["reason"] == "queue-full"
            foreground = [pool.submit("fg", str(i), revision, path, True) for i in range(8)]
            assert len(pool.pending) == 64
            assert pool.submit("fg", "overflow", revision, path, True).result()["reason"] == "queue-full"
            coalesced = pool.submit("0", "0", revision, path)
            assert futures[0].result()["reason"] == "stale"
        for future in futures[1:] + foreground + [coalesced]:
            assert future.result(10)["status"] == "accepted"
        rss = []
        for worker in pool.workers:
            limits = Path(f"/proc/{worker.child.pid}/limits").read_text()
            assert "268435456" in limits and "8388608" in limits, limits
            status = Path(f"/proc/{worker.child.pid}/status").read_text()
            rss.append(int(status.split("VmRSS:")[1].split()[0]))
        assert sum(rss) < 512 * 1024
        restarts = sum(worker.restarts for worker in pool.workers)
        pool.close()
        pool = Pool(HEAD, ledger)
        assert pool.db.execute("SELECT good FROM notes WHERE id=?", (json.dumps(("one", "note")),)).fetchone()[0] == original
        pool.close()
        return {"startup_ms": startup, "worker_rss_kib": rss, "restarts": restarts, "refusals": probes,
                "queue": 64, "reserved_foreground": 8, "last_good_restart": True}


gate("containment-last-good-queue", containment)


def fixture(index):
    # Seed/manifest: 805. Exact 90/9/1 size/depth buckets, half daily notes.
    bucket = index % 100
    size = 4096 if bucket < 90 else 32768 if bucket < 99 else 262144
    depth = 2 if bucket < 90 else 8 if bucket < 99 else 32
    # Two positive rows per note, regardless of size/depth. Managed links in half.
    source = ("---\ntitle: \"café 工作\"\ndate: 2026-10-31\n---\n"
              "- [ ] task **#area/work**\n- 09:00 event [person](p.md)\n").encode()
    source += ("  - [managed](resource.md)\n" if index % 2 == 0 else "\n").encode()
    source += ("> " * depth + "Unicode ~sub~ ^sup^ ~~strike~~ `code`\n\n").encode()
    source += b"| a | b |\n|---|---|\n| x | y |\n\n"
    # Bounded physical lines keep the fixture realistic and line-index work visible.
    padding = b"ordinary paragraph with punctuation, escapes \\* and text.\n\n"
    remaining = size - len(source)
    source += (padding * (remaining // len(padding) + 1))[:remaining]
    assert len(source) == size
    return source, size


def percentiles(values):
    values = sorted(values)
    return {str(p): values[min(len(values) - 1, (len(values) * p + 99) // 100 - 1)] for p in (50, 95, 99)}


def measure(binary, count, repeats):
    runs = []
    for run in range(repeats):
        started = time.monotonic()
        workers = [Worker(binary), Worker(binary)]
        startup_ms = (time.monotonic() - started) * 1000
        cpu_before = resource.getrusage(resource.RUSAGE_CHILDREN)
        def shard(shard):
            times = {4096: [], 32768: [], 262144: []}
            byte_count = unresolved = rows = edges = 0
            corpus_hash = hashlib.sha256()
            for index in range(shard, count, 2):
                source, size = fixture(index)
                corpus_hash.update(source)
                byte_count += size
                begin = time.monotonic_ns()
                result = workers[shard].parse(source)
                elapsed_ms = (time.monotonic_ns() - begin) / 1e6
                times[size].append(elapsed_ms)
                if result["status"] == "accepted":
                    value = result["value"]
                    assert len(value["rows"]) == 2, (index, len(value["rows"]))
                    assert len(value["relationships"]) == (1 if index % 2 == 0 else 0), index
                    rows += 2
                    edges += len(value["relationships"])
                else:
                    unresolved += 1
            return times, byte_count, unresolved, rows, edges, corpus_hash.hexdigest()
        begin = time.monotonic()
        with concurrent.futures.ThreadPoolExecutor(2) as executor:
            results = list(executor.map(shard, range(2)))
        wall = time.monotonic() - begin
        restarts = sum(worker.restarts for worker in workers)
        rss = [int(Path(f"/proc/{w.child.pid}/status").read_text().split("VmHWM:")[1].split()[0]) for w in workers]
        for worker in workers:
            worker.close()
        cpu_after = resource.getrusage(resource.RUSAGE_CHILDREN)
        buckets = {str(size): percentiles(results[0][0][size] + results[1][0][size]) for size in (4096, 32768, 262144)}
        report = {"run": run, "notes": count, "bytes": sum(r[1] for r in results),
                  "unresolved": sum(r[2] for r in results), "rows": sum(r[3] for r in results),
                  "edges": sum(r[4] for r in results), "hashes": [r[5] for r in results],
                  "startup_ms": startup_ms, "wall_s": wall, "describe_ms": buckets,
                  "worker_peak_rss_kib": rss, "restarts": restarts,
                  "worker_cpu_s": cpu_after.ru_utime + cpu_after.ru_stime - cpu_before.ru_utime - cpu_before.ru_stime}
        print("MEASURE", binary, json.dumps(report), flush=True)
        runs.append(report)
    return runs


paired = {}
for name, binary in [("baseline", BASE), ("head", HEAD)]:
    gate(name + "-10k", lambda name=name, binary=binary: paired.update({name: measure(binary, 10_000, 1)}) or "recorded")
if len(paired) == 2:
    for size in (4096, 32768, 262144):
        def per_note(size=size):
            before = paired["baseline"][0]["describe_ms"][str(size)]["95"]
            after = paired["head"][0]["describe_ms"][str(size)]["95"]
            added = after - before
            assert added <= 0.2 and after <= before * 1.1, {"baseline_ms": before, "head_ms": after, "added_ms": added, "growth": after / before, "added_budget_ms": 0.2, "growth_budget": 1.1}
            return {"baseline_ms": before, "head_ms": after, "added_ms": added}
        gate(f"combined-parser-p95-growth-{size}", per_note)

# One full 100k pass each even when a per-note gate fails. This is source-in /
# result-out parse cost only, a lower bound on the future application cold index.
for name, binary in [("baseline", BASE), ("head", HEAD)]:
    gate(name + "-100k", lambda binary=binary: measure(binary, 100_000, 1))


def incremental():
    with tempfile.TemporaryDirectory() as directory:
        path = Path(directory) / "note.md"
        source, _ = fixture(0)
        path.write_bytes(source)
        pool = Pool(HEAD, Path(directory) / "ledger.db")
        try:
            def describe(source):
                return pool.submit("one", "one", digest(source), path, True).result(5)
            assert describe(source)["status"] == "accepted"
            initial = sum(w.parses for w in pool.workers)
            for _ in range(1000):
                assert describe(source)["status"] == "accepted"
            assert sum(w.parses for w in pool.workers) == initial
            source = source.replace(b"task", b"edit", 1)
            path.write_bytes(source)
            assert describe(source)["status"] == "accepted"
            assert sum(w.parses for w in pool.workers) == initial + 1
            return {"initial": 1, "unchanged": 0, "edit": 1, "scope": "actual candidate pool revision cache; editor/application not integrated"}
        finally:
            pool.close()


gate("incremental-revision-cache", incremental)
print("SUMMARY", json.dumps({"failures": failures, "repeats": "one each; design/budget failure stops acceptance, no 3-cold/5-warm claim",
    "unproven": ["isolated extra-event/legacy-Tasks describe cost", "application ledger/DB/journal stages", "concurrent durable ack/view latency", "editor changed-region budget", "supervisor 64MiB worst-case decoded-result ceiling", "one retry per revision"]}), flush=True)
sys.exit(bool(failures))

ADR 0003 evidence migration: `docs/plans/0067-slice2/gate-pool.py` Source commit: [515559bbbb8f449d48b4487af98b8b9723842c5a](https://git.kayg.org/kayg/calternal/commit/515559bbbb8f449d48b4487af98b8b9723842c5a); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text #!/usr/bin/env python3 """Run all containment probes, then paired normal-corpus scale measurements. Run only on a locked perf VM with hash-verified release workers. No fail-fast. """ import concurrent.futures import hashlib import importlib.util import json import os from pathlib import Path import resource import signal import sqlite3 import statistics import sys import tempfile import time spec = importlib.util.spec_from_file_location("markdown_pool", sys.argv[1]) pool_module = importlib.util.module_from_spec(spec) spec.loader.exec_module(pool_module) Worker = pool_module.Worker pools = [] class Pool(pool_module.Pool): def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) pools.append(self) HEAD, BASE = sys.argv[2:4] failures = [] def gate(name, run): try: result = run() print("PASS", name, json.dumps(result), flush=True) except Exception as error: failures.append(name) print("FAIL", name, repr(error), flush=True) finally: for pool in pools: pool.close() def digest(source): return hashlib.sha256(source).hexdigest() def containment(): with tempfile.TemporaryDirectory() as directory: path, ledger = Path(directory) / "note.md", Path(directory) / "ledger.db" started = time.monotonic() pool = Pool(HEAD, ledger) startup = (time.monotonic() - started) * 1000 assert startup <= 500, startup good = b"- [ ] keep me\n- 09:00 event" path.write_bytes(good) revision = digest(good) def submit(source): path.write_bytes(source) return pool.submit("one", "note", digest(source), path, True).result(5) result = submit(good) assert result["status"] == "accepted", result assert len(result["value"]["rows"]) == 2, result original = pool.db.execute("SELECT good FROM notes").fetchone()[0] probes = {} for name, source in [("invalid-utf8", b"\xff"), ("source-limit", b"x" * (8 * 1024 * 1024)), ("depth-limit", b"> " * 256 + b"x")]: result = submit(source) assert result["status"] == "unresolved" and result["reason"] == name, result assert pool.db.execute("SELECT good FROM notes").fetchone()[0] == original assert path.read_bytes() == source probes[name] = result["reason"] for sig, expected in [(signal.SIGKILL, "worker-death"), (signal.SIGSTOP, "timeout")]: for worker in pool.workers: os.kill(worker.child.pid, sig) result = submit(good + str(sig).encode()) assert result["status"] == "unresolved" and result["reason"] == expected, result assert pool.db.execute("SELECT good FROM notes").fetchone()[0] == original probes[expected] = result["reason"] # Replace the idle killed/stopped worker too, before proving recovery. for worker in pool.workers: worker.replace() for worker in pool.workers: assert worker.parse(good)["status"] == "accepted" for fault in ["oom", "stack"]: worker = Worker(HEAD, fault=fault) try: result = worker.parse(good) assert result == pool_module.refused("worker-death"), result assert worker.parse(good)["status"] == "accepted" probes[fault] = "worker-death/replaced" finally: worker.close() # Hold dispatch under the same lock to deterministically fill/coalesce. path.write_bytes(good) with pool.condition: futures = [pool.submit(str(i % 3), str(i), revision, path) for i in range(56)] assert pool.submit("other", "over-background", revision, path).result()["reason"] == "queue-full" foreground = [pool.submit("fg", str(i), revision, path, True) for i in range(8)] assert len(pool.pending) == 64 assert pool.submit("fg", "overflow", revision, path, True).result()["reason"] == "queue-full" coalesced = pool.submit("0", "0", revision, path) assert futures[0].result()["reason"] == "stale" for future in futures[1:] + foreground + [coalesced]: assert future.result(10)["status"] == "accepted" rss = [] for worker in pool.workers: limits = Path(f"/proc/{worker.child.pid}/limits").read_text() assert "268435456" in limits and "8388608" in limits, limits status = Path(f"/proc/{worker.child.pid}/status").read_text() rss.append(int(status.split("VmRSS:")[1].split()[0])) assert sum(rss) < 512 * 1024 restarts = sum(worker.restarts for worker in pool.workers) pool.close() pool = Pool(HEAD, ledger) assert pool.db.execute("SELECT good FROM notes WHERE id=?", (json.dumps(("one", "note")),)).fetchone()[0] == original pool.close() return {"startup_ms": startup, "worker_rss_kib": rss, "restarts": restarts, "refusals": probes, "queue": 64, "reserved_foreground": 8, "last_good_restart": True} gate("containment-last-good-queue", containment) def fixture(index): # Seed/manifest: 805. Exact 90/9/1 size/depth buckets, half daily notes. bucket = index % 100 size = 4096 if bucket < 90 else 32768 if bucket < 99 else 262144 depth = 2 if bucket < 90 else 8 if bucket < 99 else 32 # Two positive rows per note, regardless of size/depth. Managed links in half. source = ("---\ntitle: \"café 工作\"\ndate: 2026-10-31\n---\n" "- [ ] task **#area/work**\n- 09:00 event [person](p.md)\n").encode() source += (" - [managed](resource.md)\n" if index % 2 == 0 else "\n").encode() source += ("> " * depth + "Unicode ~sub~ ^sup^ ~~strike~~ `code`\n\n").encode() source += b"| a | b |\n|---|---|\n| x | y |\n\n" # Bounded physical lines keep the fixture realistic and line-index work visible. padding = b"ordinary paragraph with punctuation, escapes \\* and text.\n\n" remaining = size - len(source) source += (padding * (remaining // len(padding) + 1))[:remaining] assert len(source) == size return source, size def percentiles(values): values = sorted(values) return {str(p): values[min(len(values) - 1, (len(values) * p + 99) // 100 - 1)] for p in (50, 95, 99)} def measure(binary, count, repeats): runs = [] for run in range(repeats): started = time.monotonic() workers = [Worker(binary), Worker(binary)] startup_ms = (time.monotonic() - started) * 1000 cpu_before = resource.getrusage(resource.RUSAGE_CHILDREN) def shard(shard): times = {4096: [], 32768: [], 262144: []} byte_count = unresolved = rows = edges = 0 corpus_hash = hashlib.sha256() for index in range(shard, count, 2): source, size = fixture(index) corpus_hash.update(source) byte_count += size begin = time.monotonic_ns() result = workers[shard].parse(source) elapsed_ms = (time.monotonic_ns() - begin) / 1e6 times[size].append(elapsed_ms) if result["status"] == "accepted": value = result["value"] assert len(value["rows"]) == 2, (index, len(value["rows"])) assert len(value["relationships"]) == (1 if index % 2 == 0 else 0), index rows += 2 edges += len(value["relationships"]) else: unresolved += 1 return times, byte_count, unresolved, rows, edges, corpus_hash.hexdigest() begin = time.monotonic() with concurrent.futures.ThreadPoolExecutor(2) as executor: results = list(executor.map(shard, range(2))) wall = time.monotonic() - begin restarts = sum(worker.restarts for worker in workers) rss = [int(Path(f"/proc/{w.child.pid}/status").read_text().split("VmHWM:")[1].split()[0]) for w in workers] for worker in workers: worker.close() cpu_after = resource.getrusage(resource.RUSAGE_CHILDREN) buckets = {str(size): percentiles(results[0][0][size] + results[1][0][size]) for size in (4096, 32768, 262144)} report = {"run": run, "notes": count, "bytes": sum(r[1] for r in results), "unresolved": sum(r[2] for r in results), "rows": sum(r[3] for r in results), "edges": sum(r[4] for r in results), "hashes": [r[5] for r in results], "startup_ms": startup_ms, "wall_s": wall, "describe_ms": buckets, "worker_peak_rss_kib": rss, "restarts": restarts, "worker_cpu_s": cpu_after.ru_utime + cpu_after.ru_stime - cpu_before.ru_utime - cpu_before.ru_stime} print("MEASURE", binary, json.dumps(report), flush=True) runs.append(report) return runs paired = {} for name, binary in [("baseline", BASE), ("head", HEAD)]: gate(name + "-10k", lambda name=name, binary=binary: paired.update({name: measure(binary, 10_000, 1)}) or "recorded") if len(paired) == 2: for size in (4096, 32768, 262144): def per_note(size=size): before = paired["baseline"][0]["describe_ms"][str(size)]["95"] after = paired["head"][0]["describe_ms"][str(size)]["95"] added = after - before assert added <= 0.2 and after <= before * 1.1, {"baseline_ms": before, "head_ms": after, "added_ms": added, "growth": after / before, "added_budget_ms": 0.2, "growth_budget": 1.1} return {"baseline_ms": before, "head_ms": after, "added_ms": added} gate(f"combined-parser-p95-growth-{size}", per_note) # One full 100k pass each even when a per-note gate fails. This is source-in / # result-out parse cost only, a lower bound on the future application cold index. for name, binary in [("baseline", BASE), ("head", HEAD)]: gate(name + "-100k", lambda binary=binary: measure(binary, 100_000, 1)) def incremental(): with tempfile.TemporaryDirectory() as directory: path = Path(directory) / "note.md" source, _ = fixture(0) path.write_bytes(source) pool = Pool(HEAD, Path(directory) / "ledger.db") try: def describe(source): return pool.submit("one", "one", digest(source), path, True).result(5) assert describe(source)["status"] == "accepted" initial = sum(w.parses for w in pool.workers) for _ in range(1000): assert describe(source)["status"] == "accepted" assert sum(w.parses for w in pool.workers) == initial source = source.replace(b"task", b"edit", 1) path.write_bytes(source) assert describe(source)["status"] == "accepted" assert sum(w.parses for w in pool.workers) == initial + 1 return {"initial": 1, "unchanged": 0, "edit": 1, "scope": "actual candidate pool revision cache; editor/application not integrated"} finally: pool.close() gate("incremental-revision-cache", incremental) print("SUMMARY", json.dumps({"failures": failures, "repeats": "one each; design/budget failure stops acceptance, no 3-cold/5-warm claim", "unproven": ["isolated extra-event/legacy-Tasks describe cost", "application ledger/DB/journal stages", "concurrent durable ack/view latency", "editor changed-region budget", "supervisor 64MiB worst-case decoded-result ceiling", "one retry per revision"]}), flush=True) sys.exit(bool(failures)) ```

kayg commented

2026-11-02 12:43:59 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/options-memory.py

#!/usr/bin/env python3
"""Supplemental memory probes. Run under perf-lock; no latency claims here."""
import json
from pathlib import Path
import resource
import subprocess
import sys
import time

root = Path(sys.argv[1]).resolve()


def cap():
    resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2)
    resource.setrlimit(resource.RLIMIT_CORE, (0, 0))


for size in (1024 * 1024, 2 * 1024 * 1024, 3 * 1024 * 1024, 4 * 1024 * 1024):
    child = subprocess.run([root / "in_process_containment"], input=b"\n" * size,
                           capture_output=True, timeout=3, preexec_fn=cap, check=False)
    print("MEMORY-INPUT", json.dumps({"bytes": size, "exit": child.returncode,
          "stdout": child.stdout.decode(), "stderr_tail": child.stderr[-512:].decode(errors="replace")}), flush=True)

# Linux process RSS after exec avoids the inherited Python pre-exec RSS floor
# in wait4. Sum live RSS separately; it is a quick observation, not a hard cap.
for mode in ("A", "C"):
    child = subprocess.Popen([str(root / "options-head"), mode, "10000",
                              str(root / "markdown_pool.py"), str(root / "worker-head")],
                             stdout=subprocess.DEVNULL)
    peaks = {}
    peak_sum = 0
    while child.poll() is None:
        pids = {child.pid}
        for task in Path(f"/proc/{child.pid}/task").glob("*"):
            try:
                pids.update(map(int, (task / "children").read_text().split()))
            except FileNotFoundError:
                pass
        total = 0
        for pid in pids:
            try:
                status = Path(f"/proc/{pid}/status").read_text()
                rss = int(status.split("VmRSS:")[1].split()[0])
                # Exclude the short-lived Python exec helper from worker peaks.
                name = status.split("Name:")[1].splitlines()[0].strip()
                if name not in ("python3", "python"):
                    peaks[pid] = max(peaks.get(pid, 0), rss)
                    total += rss
            except (FileNotFoundError, ProcessLookupError, IndexError):
                pass
        peak_sum = max(peak_sum, total)
        time.sleep(0.005)
    print("RSS-SAMPLE", json.dumps({"mode": mode, "exit": child.returncode,
          "peak_simultaneous_rust_rss_kib": peak_sum, "individual_peak_rss_kib": peaks,
          "interval_ms": 5, "scope": "normal 10k fixture, not worst-case bound"}), flush=True)

ADR 0003 evidence migration: `docs/plans/0067-slice2/options-memory.py` Source commit: [bda07f4763567f710ee37b30022edf610c6fece2](https://git.kayg.org/kayg/calternal/commit/bda07f4763567f710ee37b30022edf610c6fece2); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text #!/usr/bin/env python3 """Supplemental memory probes. Run under perf-lock; no latency claims here.""" import json from pathlib import Path import resource import subprocess import sys import time root = Path(sys.argv[1]).resolve() def cap(): resource.setrlimit(resource.RLIMIT_AS, (256 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_STACK, (8 * 1024 * 1024,) * 2) resource.setrlimit(resource.RLIMIT_CORE, (0, 0)) for size in (1024 * 1024, 2 * 1024 * 1024, 3 * 1024 * 1024, 4 * 1024 * 1024): child = subprocess.run([root / "in_process_containment"], input=b"\n" * size, capture_output=True, timeout=3, preexec_fn=cap, check=False) print("MEMORY-INPUT", json.dumps({"bytes": size, "exit": child.returncode, "stdout": child.stdout.decode(), "stderr_tail": child.stderr[-512:].decode(errors="replace")}), flush=True) # Linux process RSS after exec avoids the inherited Python pre-exec RSS floor # in wait4. Sum live RSS separately; it is a quick observation, not a hard cap. for mode in ("A", "C"): child = subprocess.Popen([str(root / "options-head"), mode, "10000", str(root / "markdown_pool.py"), str(root / "worker-head")], stdout=subprocess.DEVNULL) peaks = {} peak_sum = 0 while child.poll() is None: pids = {child.pid} for task in Path(f"/proc/{child.pid}/task").glob("*"): try: pids.update(map(int, (task / "children").read_text().split())) except FileNotFoundError: pass total = 0 for pid in pids: try: status = Path(f"/proc/{pid}/status").read_text() rss = int(status.split("VmRSS:")[1].split()[0]) # Exclude the short-lived Python exec helper from worker peaks. name = status.split("Name:")[1].splitlines()[0].strip() if name not in ("python3", "python"): peaks[pid] = max(peaks.get(pid, 0), rss) total += rss except (FileNotFoundError, ProcessLookupError, IndexError): pass peak_sum = max(peak_sum, total) time.sleep(0.005) print("RSS-SAMPLE", json.dumps({"mode": mode, "exit": child.returncode, "peak_simultaneous_rust_rss_kib": peak_sum, "individual_peak_rss_kib": peaks, "interval_ms": 5, "scope": "normal 10k fixture, not worst-case bound"}), flush=True) ```

kayg commented

2026-11-02 12:44:00 +00:00

Author

Owner

ADR 0003 evidence migration: docs/plans/0067-slice2/scale-manifest.json

{
  "seed": 805,
  "generator": "gate-pool.py:fixture",
  "notes": 100000,
  "bytes": 925696000,
  "buckets": [
    {"notes": 90000, "bytes_per_note": 4096, "depth": 2},
    {"notes": 9000, "bytes_per_note": 32768, "depth": 8},
    {"notes": 1000, "bytes_per_note": 262144, "depth": 32}
  ],
  "expected_task_rows": 100000,
  "expected_event_rows": 100000,
  "expected_relationships": 50000,
  "expected_unresolved": 0,
  "baseline": "9ce09e32",
  "baseline_protocol": "same framed worker shell, baseline extract unchanged, serde_json promoted from dev dependency",
  "scope": "parser/IPC lower bound; not a materialized Home, identity inventory or application cold-index measurement"
}

ADR 0003 evidence migration: `docs/plans/0067-slice2/scale-manifest.json` Source commit: [73a24063d9f7c318e3fc3f75060ab4cb0745aded](https://git.kayg.org/kayg/calternal/commit/73a24063d9f7c318e3fc3f75060ab4cb0745aded); snapshot from dev `55f35fcd`. Full original source preserved below before repository removal/stripping. Design, code and test inputs are retained where applicable. ```text { "seed": 805, "generator": "gate-pool.py:fixture", "notes": 100000, "bytes": 925696000, "buckets": [ {"notes": 90000, "bytes_per_note": 4096, "depth": 2}, {"notes": 9000, "bytes_per_note": 32768, "depth": 8}, {"notes": 1000, "bytes_per_note": 262144, "depth": 32} ], "expected_task_rows": 100000, "expected_event_rows": 100000, "expected_relationships": 50000, "expected_unresolved": 0, "baseline": "9ce09e32", "baseline_protocol": "same framed worker shell, baseline extract unchanged, serde_json promoted from dev dependency", "scope": "parser/IPC lower bound; not a materialized Home, identity inventory or application cold-index measurement" } ```

kayg referenced this pull request

2026-11-02 12:47:22 +00:00

chore: move evidence out of the repo (ADR 0003) #858

kayg referenced this pull request

2026-11-03 11:54:32 +00:00

Tasks write actions for the UI port: move, reorder, rename, version, bulk move, delete, duplicate (on the executor) #894

kayg referenced this pull request

2026-11-04 10:07:48 +00:00

Approved-but-unported UI (ship list, Nov 4) #925

kayg commented

2026-11-04 13:04:43 +00:00

Author