goutamadwant commented on PR #1127:
URL: 
https://github.com/apache/maven-compiler-plugin/pull/1127#issuecomment-5807814172

   I reproduced the stale-class failure in `mcompiler-21_class-remove` and a 
second case where a source keeps its primary class but removes a secondary 
top-level class. I've marked this PR Draft while resolving output ownership.
   
   A cleanup based on inferred source paths misses classes when the declared 
package differs from the source path. Scanning class files by `SourceFile` also 
misses `-g:none` output and can conflate same-named sources from different 
roots. The `none`/`classes` modes and a missing incremental cache leave no 
prior-source record. I don't want to ship a cleanup that can still compile 
against stale bytecode or delete another source's output.
   
   @desruisseaux, would you favor persisting exact Java output ownership across 
embedded and forked compilation, with an explicit policy for builds whose prior 
ownership is unknown?
   


-- 
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