On Fri, Oct 26, 2007 at 12:53:25PM -0400, Jesse Keating wrote: > On Fri, 26 Oct 2007 11:43:18 -0500 > "Michael Stahnke" <mastahnke@xxxxxxxxx> wrote: > > > Also keep in mind that not everyone runs rpmlint after each > > commit/change. Some only use it when submitting/reviewing new > > packages. > > Honestly we need rpmlint ran after every build in Koji as part of > post-build processing. Maybe even stuff the rpmlint output into the > koji db for future reference, or at least on the filesystem in the same > places as the buildlogs. Merely reporting rpmlint output after every build isn't going to be too scalable, because there can be alot of noise in its output. What we really need is to compare the new rpmlint output to the rpmlint output of the previous build in that build target. So that we only see 'differences' & thus can easily spot regressions. You can extend this to other RPM semantics comparisons beyond just rpmlints, so we can do better at spotting regressions in Fedora updates automatically. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
