[PATCH 1 of 2] commit: save commit message so it's not destroyed by rollback

[Martin Geisler]

Adrian Buehlmann <adrian at cadifra.com> writes:

> On 23.11.2009 00:58, Martin Geisler wrote:
>> Adrian Buehlmann <adrian at cadifra.com> writes:
>>
>>> .hg/message is perfect (without .txt)
>>
>> Not really, IMHO. Without a filename extension, my Emacs sees this as
>> a random file and wont know that it should enter text-mode.
>
> IMHO, users are not supposed to look into the .hg dir without a tool
> that knows about .hg dirs.

My editor does not need to know about .hg directories -- instead we can
convey the needed information with a simple filename suffix.

> That's why it is a hidden dir :)
>
> No other editable text files inside .hg (and further down) have a .txt
> suffix currently (e.g. .hg/patches/series).

I don't see how you can further your point by listing more example of
files that could use a proper suffix. The .hg/hgrc file ought to have
been called .hg/hgrc.ini for example, or perhaps an even more generic
name like .hg/config.ini. The "foorc" tradition probably means little to
users on Windows:

  http://www.faqs.org/docs/artu/ch10s03.html

> So tools deciding on suffix about content are a rather weak argument.

We have described the benefits of a filename suffix (easier opening on
Windows and Mac OS X, editors recognize content). Your counter-argument
is that things are already broken so let's not fix them?

> It's probably best if Matt just picks a name and ends this discussion.

I don't think we should not channel every little decision through Matt,
I'm sure he has better things to do :-)

-- 
Martin Geisler

VIFF (Virtual Ideal Functionality Framework) brings easy and efficient
SMPC (Secure Multiparty Computation) to Python. See: http://viff.dk/.