jdaugherty commented on PR #16644:
URL: https://github.com/apache/grails-core/pull/16644#issuecomment-6162586085

   Thanks, all five are in fc0ac6266f:
   
   1. The cache comment in `build.gradle` and `grails-forge/build.gradle` now 
says that, with the validator rejecting dynamic versions, the cache in practice 
covers only the snapshots; that a pull request sees the snapshot the base 
branch last resolved, for up to a day; and that a push to the base branch 
(which refetches it and saves the Gradle user home pull request runs restore) 
or `--refresh-dependencies` gets a newer one. The gradle-developer skill 
carries the same note.
   2. `BomDependencyGraph.collect` logs the configuration and project at warn 
level when resolution throws. Covering it showed that Gradle's resolution 
result is lenient: in the disabled-repository case nothing throws, the 
dependency comes back as an `UnresolvedDependencyResult` and its subtree is 
simply absent from the walk, so the check passed over it without a word. Those 
edges are warned about too, with the root cause of the failure, and each shape 
has a TestKit feature (a missing module, and a `beforeResolve` hook that 
throws).
   3. Both reports are collected and the task throws once; a feature asserts 
that a transitive range and a BOM mismatch land in the same failure.
   4. `collect` returns the walk first and `withBomVersions` adds the probed 
versions, with the probe starting from `resolvedVersions.keySet()`; 
`collectAllResolvedModuleKeys` is gone.
   5. Retitled to "No dynamic versions in any BOM-bearing configuration."
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to