On 28/01/2008, Tom Lane <tgl@xxxxxxxxxx> wrote: > "Jon Stanley" <jonstanley@xxxxxxxxx> writes: > > On Jan 27, 2008 11:26 PM, Luke Macken <lmacken@xxxxxxxxxx> wrote: > >> Sorry for the confusion. Jon's proposal[0] was approved at the last FESCo > >> meeting, but it doesn't specify what to close the bugs as. John > >> Poelstra's bug workflow[1] page illustrates Jon's proposal, but specifies > >> that bugs be closed as RAWHIDE. > > > I'm not sure what the purpose of RAWHIDE here is....it's obviously not > > rawhide by the time that it hits stable. > > One other point here, if I haven't worn out my welcome. The > previously-cited page defining bug closure states says that RAWHIDE > "should not be used for RHEL bugs", but that is obviously a RHEL-centric > definition. I argue that in the context of Fedora, RAWHIDE should only > be used to close bugs filed against the current development version (ie, > rawhide) that don't exist in any released version. If a bug has gotten > into a release branch then it should get closed as ERRATA or > CURRENTRELEASE, as appropriate. Could I draw people's attention to the fact that bz requires a version for CURRENTRELEASE closure. I _far_ prefer this over ERRATA which I regard as a "Hell, I have no idea what resolution this should have so I'll close it as such". With CURRENTRELEASE the version field indicates for other people who can replicate the same bug what version it is fixed in - dupes etc. I'd go so far as to say ERRATA should be removed completely. Cheers -- Christopher Brown http://www.chruz.com -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
