Re: Encoding of .subversion config files?

On 2006-03-03 21:22:33 +0100, Peter Lundblad wrote:
> I'm on the oposite opinion, that we should expect it to be in the
> native encoding.

On Unices at least, the encoding is related to the current process,
not to the OS, not even to a given user. For instance I personally
use different locales (either ISO-8859-1 based or UTF-8 based),
depending on the context. If you do not want to force UTF-8, then
how about something announcing the encoding early in the file,
like for XML files?

> In general, we are UTF-8 internally and natie encoding externally

and this is sometimes bad, e.g. for the file system (file names are
shared by every process, every user, and anything other than UTF-8
on some file systems will break things).

> People editing their .subversion files will do so in their native
> encoding.

No, not necessarily. Even in ISO-8859-1 locales, Emacs can edit UTF-8
files. This is something that editors can't avoid when sharing files
with other users, editing files downloaded from the net, and so on.

> The best would be if the encoding could be specified at the
> beginning of the file, but I don't know if that's feasable.

If you mean that there would be backward compatible problems, then
it could be in the comment on the first line, as a convention.

-- 
Vincent Lefèvre <vincent_at_vinc17.org> - Web: <http://www.vinc17.org/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/>
Work: CR INRIA - computer arithmetic / SPACES project at LORIA
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org