On 1/3/07, David Zeuthen <davidz@xxxxxxxxxx> wrote:
On Wed, 2007-01-03 at 15:04 -0500, Jesse Keating wrote: > On Wednesday 03 January 2007 14:52, David Zeuthen wrote: > > Maybe I'm missing something or maybe I just don't get it, but how can > > rolling back version numbers *instead* of bumping another number (Epoch) > > ever be considered a solution? Have we done this before, and if so, what > > was the justification? Thanks. > > Yes, we've rolled back before. I don't have specific examples, but it > happens. We've also flat out removed packages introduced during test times. > Introducing epochs is ugly and will hang over the package forever, and every > effort should be made to avoid it. Because we have the 'right' currently to > not worry much about going from T1 to T2 or T2 to T3 or T3 to final, we have > a way of preventing epoch. Sorry to sound non-constructive, but can we please stop breaking upgrade paths just because someone happens to think that "epochs is ugly" (left over packages is much less of a problem). I'd like to go even further and ask for our build system to enforce this rule. The justification is that it's only a number, and this practice is bordering introducing bugs by refactoring source code just because it's "less ugly" that way. Please tell me where I can file a bug (I'm not exactly sure, I used the vauge term "Build System" but is this in Bugzilla yet?) so this issue won't be forgotten. Thanks. David
This probably belongs on fedora-devel -Mike _______________________________________________ fedora-advisory-board mailing list fedora-advisory-board@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-advisory-board _______________________________________________ fedora-advisory-board-readonly mailing list fedora-advisory-board-readonly@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-advisory-board-readonly