Re: Call for testing: cvs2bzr
Greg Ward <[email protected]>
| Newsgroups | gmane.comp.version-control.subversion.cvs2svn.devel |
|---|---|
| Message-ID | <[email protected]> |
On Tue, Aug 25, 2009 at 9:26 PM, Ian Clatworthy<[email protected]> wrote: > Greg Ward wrote: > >> Ian, where are those >> build logs again? I thought you provided a link, but couldn't find it >> just now. > > https://launchpad.net/ubuntu/karmic/+source/cvs2svn/2.2.0-1 > and click on the link in the "Builds" panel on the RHS. Huh. That particular build was a total disaster (more FAIL than PASS it looks like), not just screwed up by the odd XFAIL or DeprecationWarning. But a) it's cvs2svn 2.2 and b) without run-tests -v output it's hard to know what went wrong. Hmmm: another idea for run-tests.py: when a test fails, always show its output, regardless of -v. I mean, the RunProgramException traceback is pretty useless, since it's always the same. And the output of passed tests is way too noisy. But if we could get just the output of failed tests, that would be good. [quoth Michael] > We use the testing framework from the Subversion > project (a decision, BTW, whose wisdom is becoming ever less obvious). > So we would have to make any such changes in a way that is acceptable to > them, too. Oh. Yeah. Right. I suppose this would also make it hard to implement the above idea too? Sigh. Well, whatever. I have enough on my plate what with cvs2hg, Mercurial itself, and that full-time job that I'm supposed to be doing. So I will leave run-tests.py and svntest well enough alone. Feel free to steal my either, both, or none of my ideas. ;-) Greg ------------------------------------------------------ http://cvs2svn.tigris.org/ds/viewMessage.do?dsForumId=1667&dsMessageId=2387513 To unsubscribe from this discussion, e-mail: [[email protected]].