[PATCH 4 of 4] summary: show if parents are closed heads

[Matt Mackall]

On Tue, 2010-05-11 at 11:20 +0200, Sune Foldager wrote:
> On 10-05-2010 09:59, Matt Mackall wrote:
> > On Mon, 2010-05-10 at 09:38 +0200, Gilles Moris wrote:
> >>>> # HG changeset patch
> >>>> # User Gilles Moris <gilles.moris at free.fr>
> >>>> # Date 1273438087 -7200
> >>>> # Node ID 3e17baf3a45861ee1c8e3c9f0dce46afa1570c99
> >>>> # Parent  4f227e0671ba95e075ecdc2d2a6b5f3f0072e5bf
> >>>> summary: show if parents are closed heads
> >>>
> >>> I think this ought to go on the branch: line instead.
> 
> I'm worried about that, since it will state incorrect information (or at
> least confusing) for people with >1 head per named branch. In that case,
> the branch will not be closed, but only this branch-head. It's of course
> hard to put "(current branch-head closed)" in a shorter sentence :-p.
> 
> > Summary already sticks hints in parentheses sometimes on most lines, and
> > might start sticking them elsewhere. Whereas log has had a completely
> > rigid format for the parent line for 5 years. The only thing we've ever
> > changed there is adding branch:. That was introduced at the same time as
> > named branches and so only affects folks who start using named branches,
> > while old repos and their users' tools/scripts keep working.
> 
> Analogously, people who don't use --close-branch won't be affected by
> this, right?.

Yes. The difference is that people are already using --close-branch as
it was introduced quite a while ago and are in danger of encountering
new pain from just upgrading, whereas named branches and the branch:
line in log were introduced simultaneously and thus couldn't break
existing work flows.

-- 
Mathematics is the supreme nostalgia of our time.