Front page | perl.perl5.porters |
Postings from April 2010
On Apr 27, 2010, at 5:41 AM, Ricardo Signes <perl.p5p@rjbs.manxome.org> wrote: > Jesse wrote in http://markmail.org/message/fgz6cykyjs6bejfg ... >> * Updates to dual-life modules should consist of minimal patches to >> fix crashing or security issues (as above). >> >> * New versions of dual-life modules should NOT be imported into maint. >> Those belong in the next stable series. > > I think the second point can lead to more trouble than is worth, and I'm hoping > it can be softened just a bit. I think, really, it should be viewed in terms > of the next point from that list: > >> * Patches that add or remove features are not acceptable. > > Pod::Simple had a regression in 5.12.0 that makes some Pod documentation render > incorrectly. This means that some of the core documentation can't be read > coherently, and that plenty of add-on documentation from the CPAN will also > appear incorrectly rendered. I don't think this is acceptable; maint should be > fixed so that the documentation system works. > > If we cut a new release of Pod-Simple, with only a fix to the regression, we're > not adding any features or changing any code, just fixing the serious bug. > We're not backporting a new version "just because," but to fix a bug addressed > by the new version -- which does nothing else! > > If we just cherry-pick in the changes, we need to assign a dev version number > that corresponds to nothing outside of core, and for no benefit as far as I can > tell. I think this makes sense when we've had to invent a new version because > (for example) the bugfix is only available as part of a release with additional > features or behavior changes. > > That's all. I'd like to move ahead with getting a new Pod-Simple out the door > and fixed in core, so I was just hoping for a "yes, that's reasonable" or "no, > never import a new real release to core." Thanks. That sounds reasonable and is similar to what we are doing with CGI.pm > > -- > rjbsThread Previous