Bump kompile-cli to 0.0.94 and prune transitive test dependencies - #8
Open
codexcoder21 wants to merge 1 commit into
Open
Bump kompile-cli to 0.0.94 and prune transitive test dependencies#8codexcoder21 wants to merge 1 commit into
codexcoder21 wants to merge 1 commit into
Conversation
The pin was 0.0.65, below 0.0.79 where the local build-rule result index stopped being keyed by rule name alone. Under the old scheme, editing module source did not invalidate it, so a local test run could be served a stale jar — a local green here was not evidence after a source edit or a rebase. This bit for real in PhotoGenerationManagerWui, where the local runner reported a green suite for code that was provably broken. Also ran TransitiveTestDependencyPruner: 24 @file:WithArtifact declarations across 3 of the 4 test files, all supplied transitively by the module artifact already. testEvergreenServiceServerEndToEnd.kts is deliberately untouched. The pruner could not locate the compile dependencies of the buildImageClientResourcesJar() build rule in build.kts, so what that rule provides transitively is unknown and no removal there could be proven safe. It reports the skip rather than guessing, and that judgement is left standing rather than forced. 64/64 on the full suite remotely. Co-Authored-By: Claude Opus 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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Two pieces of build/test hygiene.
kompile-cli 0.0.65 → 0.0.94
The pin was below 0.0.79, where the local build-rule result index stopped being keyed by rule name alone. Under the old scheme, editing module source did not invalidate the cached artifact, so a local
scripts/test.bashrun could be served a stale jar — meaning a local green in this repo was not evidence of anything after a source edit or a rebase.Not hypothetical: in PhotoGenerationManagerWui the local runner reported
ALL TESTS PASSED (5/5)against code that had been verified defective, and the same tree on 0.0.94 correctly failed. Only the pinned version differed. 89% of the org is still below the threshold — challenge.Pruned transitive test dependencies
Ran TransitiveTestDependencyPruner: 24
@file:WithArtifactdeclarations across 3 of the 4 test files — coordinates the module artifact already supplies transitively, left over from when@WithArtifactwas not transitive.tests/testEvergreenServiceServerEndToEnd.ktsis deliberately left alone. The pruner could not locate the compile dependencies of thebuildImageClientResourcesJar()build rule inbuild.kts, so the set of transitively-provided dependencies for that file is unknown and no removal could be proven safe:That is the tool declining to guess, and the right call — so this repo gets a deliberately partial clean rather than a forced complete one. Whatever redundancy that file carries stays until the build rule is expressed in a way the pruner can read.
Why one PR
Both changes are build/test hygiene on the same repo and are proven by the same suite run; splitting them would cost the shared buildtest droplet two extra CI rounds for no added signal — a wrongly-pruned coordinate surfaces as a test file that will not compile either way.
64/64 on
scripts/test.bash --remote --test ..🤖 Generated with Claude Code