Martin: On Tue, Jul 28, 2009 at 1:57 PM, Martin J. Evans<martin.ev...@easysoft.com> wrote: > Jonathan Yu wrote: >> Hi: >> >> I seek the wisdom of any other module authors that might have come >> across this problem. >> >> Recently, I uploaded a new version of Math::Random::ISAAC::XS and ran >> into a *lot* of regressions. I've looked at the diff and I didn't >> really change all that much, except for removing some things from >> Recommends. The smokers nonetheless output something that I can't >> reproduce, and I'm not sure if it has to do with my use of the >> 'version' pragma, or if the systems in question are using an older >> version than I test with. >> >> In at least one report, the version seems to be recent, so I'm not >> sure if it's a new issue: version 0 0.76_06 >> >> I get plenty of failing tests: >> http://www.cpantesters.org/distro/M/Math-Random-ISAAC-XS.html#Math-Random-ISAAC-XS-1.0.6 >> >> Here is the diff between my last (100% PASS) version, 1.05, and the >> latest version, which has a lot more failures than I'd like: >> http://search.cpan.org/diff?from=Math-Random-ISAAC-XS-1.0.5&to=Math-Random-ISAAC-XS-1.0.6 >> >> Any insight that the module-authors can provide would be greatly >> appreciated. Does this have to do with the latest version pragma? >> Maybe I should also agree that it's considered a Bad Thing and move to >> using the older, more Perlish version numbers :( >> >> Cheers, >> >> Jonathan >> >> > > I think a new version of version.pm was made available a few days ago by > David Golden (0.77). You fails results seem to mention if (0.77). > Perhaps this is the cause. I've cc'ed David just in case (apologies if > I'm off the mark on this David).
On further analysis I don't think version 0.77 is the culprit -- I've installed that version inside a clean chroot and everything still builds per normal. I'm not sure if it was something related to that version somehow though. I suppose part of the issue is that CPAN Smokers don't provide a full build log, but only a report of the test part, so that might be hampering the investigation. And of course, until I can figure out how to reproduce the bug, I can't fix it > > Martin >
