Repository navigation
Fix master CI: Flatpak archive type and two stale Windows tests - #99
Closed
carrdylan52 wants to merge 4 commits into
Closed
carrdylan52 wants to merge 4 commits into
carrdylan52 wants to merge 4 commits into
Conversation
flatpak-builder picks an archive's format from its file name, and the codeload URLs end in the bare tag (v0.9.1, v4.6.11), so the Flatpak job has stopped with "Unknown archive format" since 20fc7f6 added them. Both are gzip tarballs; the pinned digests are unchanged.
6d99d9f (TheMoonPeople#68) stopped decoding every variant of a named picture at registration; the alternates now decode when selected. The test still counted two decodes per entry, so it has failed on Windows CI since TheMoonPeople#68. Count one decode for named entries and check that selecting the alternate decodes it then.
datapacks::Load canonicalises the data root, which expands the Windows runner's short (8.3) TEMP name (RUNNER~1 to runneradmin), so the expected paths built from temp_directory_path() never matched on CI. Build the fixture root from the canonical temp directory.
Since TheMoonPeople#68 a fully named picture decodes its other variants when they are selected, and a changed identity replaces the picture rather than keeping it by its pixels. The MgTextureEnterNativeSet comment and the test's comment said otherwise.
Adubbz
pushed a commit
that referenced
this pull request
Oct 18, 2026
Name the archive type of the fastgltf and simdjson tarballs, whose URLs end in a bare tag. Expect named pictures to decode their other variants when selected, compare datapack candidates against the canonical temp root, and remove the whole legacy card directory before replacing it with a file. Co-Authored-By: millerscout <amarcosju@gmail.com> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Before: every master CI job is red. Flatpak stops at
Unknown archive formaton the fastgltf download; Windows failsMgTexture.NativeSetsReuseIdenticallyNamedPicturesWithoutDecodingandNativeAssetsTest.DirectAssetCandidates...; Linux and macOS stop at configure on a stale prebuilt FFmpeg (fixed in Kazarov, see below).After: the Flatpak and both Windows failures are fixed here.
How:
archive-type: tar-gzip(hashes unchanged).%TEMP%is an 8.3 path (RUNNER~1) thatdatapacks.cppexpands to the long form, while the test built its expected paths from the short one. The test root is now the canonical temp path.Tests:
ctest -LE gpupasses on Windows with an 8.3%TEMP%.Still needed for Linux/macOS: Kazarov's prebuilt fix (TheMoonPeople/Kazarov PR), then a follow-up here pointing
tools/ci/prebuilt-deps.shat the new release with its SHA256SUMS. The macOS runner's SDK also had atapi ... arm64e.x1link error before configure broke, which may resurface.