[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Deprecation of svn_client_resolve() leaves loose ends

From: Daniel Shahaf <d.s_at_daniel.shahaf.name>
Date: Thu, 21 Apr 2016 10:36:17 +0000

Branko Čibej wrote on Sun, Apr 17, 2016 at 18:47:35 +0200:
> P.S.: It would also be nice if developers compiled stuff in maintainer
> mode and fixed new warnings before committing ... just sayin'.

That's exactly why we have the svn-warnings bot. Its build did trigger
the 'switch' warning you had just fixed elsewhere, but that didn't cause
the build status to become red/failure:

https://ci.apache.org/builders/svn-warnings/builds/1145/steps/Make/logs/stdio

So, I guess our custom "fail the build if stderr is non-empty" builder
setup code needs some attention:

https://svn.apache.org/repos/infra/infrastructure/buildbot/aegis/buildmaster/master1/projects/subversion.conf
[committers-only link, sorry]

Cheers,

Daniel

P.S. The existing bot uses gcc. We could add a clang warnings bot if we
wanted to, to improve coverage... (it just runs 'make all install'
without 'make test', so requires little resources on the buildslave)
Received on 2016-04-21 12:36:19 CEST

This is an archived mail posted to the Subversion Dev mailing list.