Re: Re* Regulator updates for 3.3

From: Linus Torvalds <torvalds@linux-foundation.org>
To: Junio C Hamano <gitster@pobox.com>
Cc: Mark Brown <broonie@opensource.wolfsonmicro.com>,
	Liam Girdwood <lrg@ti.com>,
	linux-kernel@vger.kernel.org,
	Git Mailing List <git@vger.kernel.org>
Subject: Re: Re* Regulator updates for 3.3
Date: Wed, 11 Jan 2012 08:23:01 -0800	[thread overview]
Message-ID: <CA+55aFy679Skqi_D3x8=M=mwZiViMX9EbZrqP11riiLb_Hzb9g@mail.gmail.com> (raw)
In-Reply-To: <7vzkdu7miv.fsf@alter.siamese.dyndns.org>

On Tue, Jan 10, 2012 at 10:59 PM, Junio C Hamano <gitster@pobox.com> wrote:
>
> What makes me uneasy about the idea of running the editor by default is
> that many people still use Git as a better CVS/SVN. Their workflow is to
> build randomly on their 'master', attempt to push and get rejected, pull
> only so that they can push out, and then push the merge result out.

Sure. And I don't think we can do much about it. They'll either set
the legacy flag, or they'll just exit the editor without adding
anything useful (if you come from a CVS background in particular, you
probably never learnt to do good commit logs anyway).

So it will be a bit more work for the bad workflow, I agree - although
if it really irritates people, they can just set that GIT_MERGE_LEGACY
in their .bashrc files or something. But we can *hope* that even those
people might sometimes actually talk about what/why they are doing
things, or maybe even learn about that whole "distributed" thing.

I agree that is unlikely to ever happen, though. It's more likely that
they will change their aliases so that their "update" command just
adds the --no-edit flag. Regardless, it doesn't sound *too* onerous to
work around.

Patch looks good to me. I would personally have compared "st_mode"
instead of (or in addition to) "st_rdev", but I don't think it matters
all that much.

                                 Linus