On Mon, May 06, 2019 at 04:44:43PM +0100, Ian Jackson wrote:
> Colin Watson writes ("Bug#928473: dgit: not clear what to do when earlier
> uploads used dgit but intermediate ones didn't"):
> > I'd been hoping to avoid the pseudomerge and have the histories be
> > exactly identical, but I probably shouldn't be too precious about that,
> > and you make a good point about allowing people with existing dgit
> > clones to fast-forward.
>
> Other than in split-brain quilt modes (which don't apply in this
> case), the pseudomerge will appears on the branch you run `dgit push'
> from. So the histories will be identical because the pseudomerge will
> be in your master. Is that a problem ? I thought I should mention
> it...
Oh, so it did. Well, it's too late now in any case, but I think on
balance I prefer having it there in both master and dgit/dgit/sid rather
than just the latter. Thanks for the note.
--
Colin Watson [cjwat...@debian.org]
