Re: What's cooking in git.git (Jun 2016, #04; Tue, 14)

From: Mike Hommey <mh@glandium.org>
Date: Jun 16, 2016, 04:16 UTC
Message-ID: <20160616041630.GA32045@glandium.org>
In-Reply-To: <xmqqd1ni716g.fsf@gitster.mtv.corp.google.com>

On Wed, Jun 15, 2016 at 11:32:39AM -0700, Junio C Hamano wrote:

Show 27 quoted lines

> Mike Hommey <mh@glandium.org> writes:
> 
> > On Tue, Jun 14, 2016 at 03:08:04PM -0700, Junio C Hamano wrote:
> 
> >> * mh/connect (2016-06-06) 10 commits
> >>  - connect: [host:port] is legacy for ssh
> >> ...
> >>  - connect: document why we sometimes call get_port after get_host_and_port
> >> 
> >>  Ok, folks, is everybody happy with this version?
> >
> > $gmane/296609
> > $gmane/296610
> 
> Oh, I have seen these, and I know you two are happy.
> 
> But I am having a hard time coming up with a few-line summary for
> this topic.  I can write the beginning part, i.e. "Git-URL parsing
> routine has been rewritten", but the concluding part of the sentence
> cannot be "... has been rewritten for no good reason." if I were to
> mark the topic as "Will merge to 'next'".  The best I can come up
> with is "... has been rewritten (hopefully) without changing the
> benaviour.", but that is not a strong-enough justificaiton to make
> the change to the codebase, either.
> 
> In short, while the update may not introduce new bugs, why would we
> want to have this change in the first place?

My original motivation was to avoid having to copy code from connect.c into git-cinnabar, which is what I'm currently doing[1]. Things derailed a little, and we got ourselves somewhat in the middle of a refactor, that I'm willing to push a little further (like, refactor things such that host_end only happens once). My hope is that this makes the code more maintainable.

Mike

1. https://github.com/glandium/git-cinnabar/blob/master/connect.c