release: v1.1.0 #70

Merged

yuris merged 1 commit from release/v1.1.0 into main

2026-08-20 16:29:23 +00:00

yuris commented

2026-08-20 16:29:16 +00:00

Owner

Final Phase 5 item of the v1.1.0 milestone (#59).

game.project's version goes 1.0.0 → 1.1.0, and the three v1.0.0 URLs in
README.md become v1.1.0: the archive link in the Install block, the
private-repo token example, and the HTML5 patch-fetch.

The tag those URLs name does not exist until this merges, which is fine —
html5-conformance.sh applies patches/ from the working tree, not from the URL,
so conformance does not depend on it.

Verification

lua dev/test/runner.lua       7 passed, 0 failed
conformance.sh                PASS, PHASE1-6, identity persisted across a restart
html5-conformance.sh          PASS

The archive gate flagged this bump as the one blocking item left, and noted the
archive verification should be re-run against the tag's own artifact afterwards,
since this changes two shipped members (game.project, README.md). The file
list is unaffected, so the composition, .gitattributes coverage and
require-resolution results carry over.

One harness observation, not a blocker

html5-conformance.sh returned FAIL once on this branch and then passed four
times in a row (28/29/28s against a 120s timeout). I did not capture the frames,
so I cannot attribute it; recorded on #52, which exists for exactly this and asks
for a frame capture next time. Explicitly ruled out: it is not PHASE6 pushing the
per-load budget, since the passing runs sit at ~4× headroom and a timeout emits
its own distinct message. dev/ is export-ignored, so this is harness
reliability, not shipped behaviour — and it is the standing reason #16's
continue-on-error stays on the HTML5 CI step.

Final Phase 5 item of the v1.1.0 milestone (#59). `game.project`'s `version` goes `1.0.0` → `1.1.0`, and the three `v1.0.0` URLs in `README.md` become `v1.1.0`: the archive link in the Install block, the private-repo token example, and the HTML5 patch-fetch. The tag those URLs name does not exist until this merges, which is fine — `html5-conformance.sh` applies `patches/` from the working tree, not from the URL, so conformance does not depend on it. ## Verification ``` lua dev/test/runner.lua 7 passed, 0 failed conformance.sh PASS, PHASE1-6, identity persisted across a restart html5-conformance.sh PASS ``` The archive gate flagged this bump as the one blocking item left, and noted the archive verification should be re-run against the tag's own artifact afterwards, since this changes two shipped members (`game.project`, `README.md`). The file *list* is unaffected, so the composition, `.gitattributes` coverage and require-resolution results carry over. ## One harness observation, not a blocker `html5-conformance.sh` returned `FAIL` once on this branch and then passed four times in a row (28/29/28s against a 120s timeout). I did not capture the frames, so I cannot attribute it; recorded on #52, which exists for exactly this and asks for a frame capture next time. Explicitly ruled out: it is not PHASE6 pushing the per-load budget, since the passing runs sit at ~4× headroom and a timeout emits its own distinct message. `dev/` is export-ignored, so this is harness reliability, not shipped behaviour — and it is the standing reason #16's `continue-on-error` stays on the HTML5 CI step.

yuris added 1 commit

2026-08-20 16:29:16 +00:00

release: v1.1.0

ci / ci (pull_request) Successful in 19s

Details

1120771995

Bumps game.project's version and points the three install URLs at the v1.1.0
tag: the archive link, the private-repo token example, and the HTML5
patch-fetch. The harness applies patches/ from the working tree rather than the
URL, so the tag not existing yet does not break conformance.

yuris merged commit a60640662a into main

2026-08-20 16:29:23 +00:00

yuris referenced this pull request from a commit

2026-08-20 16:29:25 +00:00

release: v1.1.0 (#70)