Re: [PATCH (GIT-GUI,GITK) 1/8] git-gui: Cleanup handling of the default encoding.

["Alexander Gavrilov" <angavrilov@gmail.com>]

On Thu, Sep 18, 2008 at 8:50 PM, Dmitry Potapov <dpotapov@gmail.com> wrote:
> It is not about how data are stored locale but what is in repository.
> Even if you still have some Linux box with legacy encoding on it, you
> still want to see what in repository, which is mostly likely to be in
> UTF-8. Even if you do not have UTF-8 locale, all decent editors are
> capable to read and store files in UTF-8 (even if it is not your locale),
> and it is really make sense to store files in UTF-8, which makes sense
> because you are going then on a modern Linux, you want to have all data
> in the repository to be in a single encoding, and UTF-8 is the best
> choice for that.

A new user would expect to see his files properly, and they are likely
to be in the locale encoding. And if you know about utf-8, you can
open the Options dialog, and select it explicitly from a menu. And if
you commit a .gitattributes file with encoding specifications to the
repository, it will be used automatically wherever you check it out.

> This patch is certainly a big improvement, as it allows to choose what
> encoding you want to see, but I was not sure that changing the default
> from UTF-8 to the system locale is really a good idea for anything but
> Windows specific projects. Anyway, I have converted all computers that
> I use regularly to UTF-8, so I don't really care...

You are here missing the fact, that the actual current default for
git-gui is not utf-8, but 'binary', essentially equivalent to
ISO-8859-1. UTF-8 was suggested by a patch that has been around in the
'pu' branch since January, and which I took as a base for my series.
Gitk on the other hand uses the locale encoding.

Alexander