[
https://issues.apache.org/jira/browse/LUCENE-5467?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Uwe Schindler resolved LUCENE-5467.
-----------------------------------
Resolution: Fixed
> Re-add "ant clean-jars" on root level
> -------------------------------------
>
> Key: LUCENE-5467
> URL: https://issues.apache.org/jira/browse/LUCENE-5467
> Project: Lucene - Core
> Issue Type: Improvement
> Components: general/build
> Affects Versions: 4.7
> Reporter: Uwe Schindler
> Assignee: Uwe Schindler
> Fix For: 4.8, 5.0
>
> Attachments: LUCENE-5467.patch
>
>
> Although {{ivy sync=true}} seems to work in most cases, sometimes (especially
> when updating version numbers) the old JAR files are not nuked on resolve. I
> have no idea about why this is still a bug in ivy, but sync=true does not
> help in all cases.
> It would be good to re-add the {{ant clean-jars}} target to the top-level
> build.xml, because this allows you to remove all JAR files and do a
> re-resolve.
> The bug occurred to [[email protected]] and also to me this morning when
> Mark updated slf4j versions: If you run {{ant jar-checksums}} after updating
> the version numbers, it creates 2 checksum files (for the old and the new
> version). The call to {{ant check-svn-working-copy}} then does not find a
> violation, because the old version's checksum is still recreated. It only
> complains about the new one, but when you add it it will not complain anymore
> about the old one.
> The violation is only found by Jenkins, because this one does a fresh svn
> checkout (or it emulates one), so it starts without any JARs. Because of
> this, checksum re-creation deletes the old file (still on subversion server)
> and a violation is detected.
> Currently I do the JAR file deletion by a find/xargs/rm and then recreate
> checksums, which finds the problem. The good old {{ant clean-jars}} would be
> back as a last resort if you are hit my this IVY bug when you are updating
> revision numbers.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
