[PATCH] debugbundle: handle the --all option for bundle2

[Mike Hommey]

On Wed, Apr 20, 2016 at 05:53:18PM -0500, Matt Mackall wrote:
> On Thu, 2016-04-21 at 06:50 +0900, Mike Hommey wrote:
> > On Wed, Apr 20, 2016 at 01:40:40PM -0700, Pierre-Yves David wrote:
> > > 
> > > Thanks for your patch, I really appreciate how you quickly tackle issue your
> > > find.
> > > 
> > > However, we are currently in feature freeze for 3.8 and this patches seems
> > > large enough
> > It's only large because of the indentation level change.
> > 
> > > 
> > > that we would not take such churn for the sake of a debug
> > > command. Can you resent it after 3.8 is release (May first) for review?
> > Sure, but I'll take on the occasion to ask why request of patch authors
> > to resend instead of keeping a queue of those patches in some form
> > (actual patch queue, mercurial branch, whatever).
> > 
> > > 
> > > (also, your patch seems to be missing the expected metadata header and
> > > decription, how did you sent it?)
> > git-send-email. What metadata header are you expecting in particular?
> > User and Date? Do you care about Node-ID and Parent?
> 
> With the changeset identifier, hg import can add an appropriate obsolete marker.

But that's only valuable to the patch sender, right? Does the Parent
have a value for you or can I get away with only User, Date and
repeating the patch summary below the header?

Mike