Skip to content

ci: extend SignPath signing wait timeout to 1 hour - #1052

Merged
PCMan merged 1 commit into
masterfrom
ci/signpath-timeout
Nov 21, 2026
Merged

PCMan merged 1 commit into
masterfrom
ci/signpath-timeout

Conversation

@PCMan

@PCMan PCMan commented Nov 21, 2026

Copy link
Copy Markdown
Member

Set \wait-for-completion-timeout-in-seconds: 3600\ on both the \Sign binaries\ and \Sign installer\ steps in .github/workflows/build-and-sign.yaml.

The default 600-second (10-minute) timeout is frequently too short when waiting for manual SignPath release-signing approval, causing release workflows to time out mid-flight.

Fully generated by Google Antigravity (Gemini 4 Neon) without human intervention.

Set wait-for-completion-timeout-in-seconds to 3600 on both Sign binaries and Sign installer steps in .github/workflows/build-and-sign.yaml.

The default 600-second (10-minute) timeout is too short for manual SignPath release-signing approvals and caused release builds to fail while waiting for maintainer sign-off.

Fully generated by Google Antigravity (Gemini 4 Neon) without human intervention.
@PCMan
PCMan merged commit 1911328 into master Nov 21, 2026
6 of 8 checks passed
@PCMan
PCMan deleted the ci/signpath-timeout branch November 21, 2026 10:46
asdf80730 added a commit to asdf80730/PIME that referenced this pull request Nov 22, 2026
Upstream cut v2.0.0-beta6 at 1911328 on 2026-11-21: EasyIME#1043 config-tool token
check, EasyIME#1049 the cinbase Shift key-up path (EasyIME#1044), EasyIME#1009 and EasyIME#1050 the Go
backend imports and optional build, EasyIME#1051 the Rime auto-deploy, EasyIME#1052 the
SignPath wait, and the version.txt bump. No feat(...) landed, so the 2026-11-13
"v2.0 first" adjudication still holds.

Conflict resolutions:

- go-backend/server.go keeps the local side, which is a superset: the guids
  loop and the "rime", "rime-cn" case (ADR 0004) over upstream's single guid.
- go-backend/server_integration_test.go takes upstream's close-notification test.
- python/cinbase/__init__.py takes upstream's lastKeyEvent key-up path, with the
  local shiftTapCount loop kept so N taps still toggle N times; the local
  shiftTapMatches now passes cbTS to isPressed for its per-side check.
- python/input_methods/rime/rime_ime.py stays the thin wrapper; upstream's
  needs_rime_deploy and fullcheck land in python/rimebase/rime_service.py and
  are re-exported by both profile wrappers, and tests/test_rime_deploy.py is
  re-pointed at that module pair.

docs/tour/index.html and the maintainer-plan record now read against 1911328.

Fully generated by omp (Qwen3.8-Flash-Next) without human intervention.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant