Re: Branching off 2.2

On Tue, 2004-12-28 at 21:46 -0500, Mark Gilbert wrote:

> > I never had any problems with 1. But I wouldn't be against 2 either. I
> > think 1 is simpler for developers, as they then only need to have 1
> > branch (HEAD).

How having 2 branches is something complicated ?
BTW what we want is make our user's life easier. By using solution #2,
this is just what we do because we fix bug in the current release...

> I never had any problems with 1 and I never argued against 2, what I
> argued against was sweeping judgements about all fixes needing to go one
> way or the other. For minor, unintrusive, unlikely-regressive, and
> same-for-both-lines fixes, any of 1, 2, and 3 works, the important part
> being that the fix happens and not where the acting coder finds it most
> convenient. For intrusive fixes, possibly regressive fixes, and
> anything that requires divergent infrastructure in HEAD, 1 is IMO the
> only option. Using 3 for trivial and no-risk fixes means less
> administrative overhead for the maintainer, but that administrative
> overhead really is negligible compared to the fix happening _somewhere_,
> and being ported where applicable.

So you declare willing to be repsonsible off backporting everything ?
I'm sorry, but I have done enough backporting do declare it to be a PITA
process, not including time consuming.

Hub

-- 
Crazy French - http://www.figuiere.net/hub/