Front page | perl.perl5.porters |
Postings from July 2010
On Sun, Jul 11, 2010 at 9:06 PM, David Golden <xdaveg@gmail.com> wrote: > So I pose this question to Jesse and the rest of p5p -- for major > changes, do we want it in blead? Do we want it kept in feature > branches until it passes muster? > > What do people think? I think the chances of getting it to pass muster without getting it into blead first are pretty much nil. In the past, feature branches, however ready people thought they were, have tended to have plenty of unpleasant surprises once they got into blead. We recently had a simple, innocent-looking integration of a CPAN module that broke the build on three or four platforms. So I really don't think we even have a way to know whether something passes muster if it's not in blead. I can understand that the monthly release managers might be wary of an ambitious patch that could lead to black smoke near the time of their release, and I appreciate their efforts to keep things as close to release-ready as possible. But I think the only thing we can do without the expenditure of more (unavailable) people time is to either move the time of the release or move the time that the larger, riskier patch gets applied. In the meantime anyone can test a patch (even if he or she doesn't understand what it's doing) and add a note to the ticket of how cleanly it applied and whether it passed tests.Thread Previous | Thread Next