Quoth Danek Duvall on Wed, Aug 20, 2008 at 04:12:55PM -0700: > On Wed, Aug 20, 2008 at 03:34:48PM -0700, David Bustos wrote: > > Aside: Speaking of which, I've been wondering if we should file > > an RFE with Mercurial for an option which pulls not the > > changeset for a tag, but the changeset which *defines* the tag. > > Like > > > > hg pull -R onnv_96 > > > > Except, of course, that -R is already used. Maybe some sort of > > defined:onnv_96 syntax could work instead. > > The problem with this is that it requires that the changeset that > introduces the tag come immediately after the changeset it tags, and that > isn't always going to be the case. It could come in after an arbitrary > number of changesets, possibly out of order with other tags, etc, and so > you end up with a bunch of changes that are explicitly not supposed to be > part of the tree you're testing.
Yes, that's true. Shoot. > Now, if ON always gets it right, maybe such an RFE would be useful. But I > don't see project gates sticking to what's actually a fairly strict rule, > so I'm not sure it'd be all that useful in the end. Do you mean you think project gatekeepers are more apt to make changes between an O/N tag changeset and the changeset it labels? That's the point of the RFE -- to make it easy for project gatekeepers to maintain the rule that O/N does. Though even if there are changes inbetween, wouldn't they be on a different branch than the tag changeset? Maybe Cadmium can look at the next changeset, check whether it affects anything other than .hgtags, and pull it in. Eh, probably not worth it. David
