sorokin.oleg.edu
36db9356cc
Preserve loaded draft bytes and model history, reject stale workspace and input events, and keep captured Escape under modal priority. Verified 181 focused Windows tests and strict desktop Clippy on exact inputs4; earlier full737 scope and component/native preparation evidence are archived separately. Native navigation acceptance remains pending.
1.4 KiB
Source navigation follow-up gates
Codex-9e8r6t /root, Windows 2026-09-11. These exact records complement the initial desktop gates and the integrated evaluation. Linux/WSL remains user-paused.
inputs3 covers ordinary Escape cleanup; focused179 and strict all-target
Clippy passed, but review then found the captured-modal-input activation gap.
inputs4 extends the single captured-Escape bridge under ordinary Text modals,
avoids the separate architecture route, and tests the real App subscription
using a captured event from a mounted TextInput. Its terminal results are in
focused4 and clippy4. Both raw source snapshots read back unchanged.
The original GUI build passed from inputs2 and was never launched. Its preserved Cargo JSON and receipt identify binary9FA1457E...53CB66; that binary is omitted and superseded for native acceptance. The helper is preserved unchanged with its original private paths. These records do not establish a newer GUI/native pass.
Reviews retain their original findings and exact source hashes: review.md is
against inputs2, review-escape.md against inputs3. Later corrections do not turn
those earlier findings into passes. Any later source review has its own scope.
manifest.json indexes all payload bytes including this README and attributes;
readback.json pins the manifest. Root verifies raw Git-index bytes on staging.