On Tue, 2006-01-03 at 19:07 +0100, Arjan van de Ven wrote: > > > > > So we make its life simple, and stick with a 2.6.x-CVSident_$releasever > > > > I disagree. IMHO is confuses users, journalists (they wrote "FC4 is > > there is one angle that davej didn't mention yet: the version is named > after the version of the tarbal. The git etc patches are that. patches > on top of the tarbal. But the tarbal version still is 2.6.14... > btw if I had a chance to do things again.. I'd use the version "2.6" without any subrelease number. And just leave the "14" or "15" out of the rpm versioning entirely. Too much emphasis is put on that exact subversion number anyway while the meaning of it is rather, ehm, minute. Especially since a lot of bugfixes of later kernels get backported. People still say "but fedora has THAT bug because it has kernel version 2.6.14, see the rpm says so" even when the fix for THAT bug has been in the rpm for forever. Just leaving this out (and using the auto incrementing cvs id) sounds the best to me to be honest. Or maybe 2.6.FC4, 2.6.FC5 etc -- fedora-test-list mailing list fedora-test-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list