S1 vt: snapshot fidelity: REP, cursor and line-edge mismatches after replay #25

New issue

Closed

opened 2026-10-06 15:42:27 +00:00 by kayg · 2 comments

kayg commented

2026-10-06 15:42:27 +00:00

Owner

Part of #22. The independent review of the libghostty-vt switch (head d3c7194) returned rework before it was cut short by the AI tooling's content filter:

F1 (high): Snapshot/replay loses REP state. The alternate screen is painted after the primary in crates/infraternal-shim/native/snapshot.zig:40–48.
A cursor-position mismatch after replay.
Line-edge (wide or continuation) cell mismatches.

Job: fix-vt (Codex GPT 6.1 Sol Medium), worktree .worktrees/vt-ghostty.
It builds a differential snapshot test (source terminal vs replayed terminal: grid, cursor, modes, last printed char, pending wrap) with a seeded 20k-case corpus, then fixes each mismatch.

Done when: the differential suite passes, the flood budget of 10 MB/s still holds, the fast gate is green, and a fresh review (FFI soundness, carry observer, licences) approves.

Part of #22. The independent review of the libghostty-vt switch (head d3c7194) returned **rework** before it was cut short by the AI tooling's content filter: - **F1 (high):** Snapshot/replay loses REP state. The alternate screen is painted after the primary in `crates/infraternal-shim/native/snapshot.zig:40–48`. - A cursor-position mismatch after replay. - Line-edge (wide or continuation) cell mismatches. **Job:** fix-vt (Codex GPT 6.1 Sol Medium), worktree `.worktrees/vt-ghostty`. It builds a differential snapshot test (source terminal vs replayed terminal: grid, cursor, modes, last printed char, pending wrap) with a seeded 20k-case corpus, then fixes each mismatch. **Done when:** the differential suite passes, the flood budget of 10 MB/s still holds, the fast gate is green, and a fresh review (FFI soundness, carry observer, licences) approves.

kayg referenced this issue

2026-10-06 15:42:51 +00:00

Status board: work in flight (kept current) #24

kayg referenced this issue

2026-10-06 15:42:52 +00:00

S1: switch the shim to libghostty-vt #22

kayg commented

2026-10-06 17:31:02 +00:00

Author

Owner

fix-vt finished (vt-ghostty 234baf0, 21 commits): screen ordering, REP and pending-wrap restore, cursor, line-edge metadata. Nightly: 19,864 faithful replays, 136 explicit API-gap rejections (Unsupported), 0 accepted divergences. Flood is still above 10 MB/s. Snapshot p99 is about 1.0–1.2 ms (4.6 ms on a loaded host). Upstream setter proposals are in docs/S1-VT-GHOSTTY.md. The second review runs as #26.

fix-vt finished (vt-ghostty 234baf0, 21 commits): screen ordering, REP and pending-wrap restore, cursor, line-edge metadata. Nightly: 19,864 faithful replays, 136 explicit API-gap rejections (Unsupported), 0 accepted divergences. Flood is still above 10 MB/s. Snapshot p99 is about 1.0–1.2 ms (4.6 ms on a loaded host). Upstream setter proposals are in docs/S1-VT-GHOSTTY.md. The second review runs as #26.

kayg referenced this issue

2026-10-06 17:31:03 +00:00

Status board: work in flight (kept current) #24

kayg commented

2026-10-28 17:52:15 +00:00

Author

Owner

Closed by the orchestrator during the 2026-10-28 handover (#254), on the owner's request to close finished issues. Evidence: fix-vt snapshot fidelity finished at 234baf0 (in dev and prod); nightly 19,864 replays, 0 divergences; second review was #26. Reopen if anything is still missing.

Closed by the orchestrator during the 2026-10-28 handover (#254), on the owner's request to close finished issues. Evidence: fix-vt snapshot fidelity finished at 234baf0 (in dev and prod); nightly 19,864 replays, 0 divergences; second review was #26. Reopen if anything is still missing.

kayg closed this issue

2026-10-28 17:52:16 +00:00

~~kayg referenced this issue 2026-10-28 17:53:54 +00:00~~

Orchestrator handover 2026-10-28: moving to a new machine #254

kayg referenced this issue

2026-10-28 18:00:45 +00:00

Orchestrator handover 2026-10-28: moving to a new machine #254