Re: problems with i18n testsuite

graydon hoare <graydon@xxxxxxxxx> writes:

> the new versions of the autotest macros reset all the locale-related
> environment variables to "C" (ascii) before running tests. this is
> just wrong imho, but I have put in work-around code for the time being.

I guess the reason is that usually (i.e. when you're not testing i18n
itself), you want your test result to be predictable, and not
influenced by the user's locale. But of course there should be a way
around this cleverness.

>> by the way: shouldn't filenames always be encoded in utf-8?
>> at least many file managers seem to use utf-8 file names.

But there are a number filesystems still containing legacy encodings.
Mine, for example, will be in latin1 until zsh and xemacs support
UTF-8 well.

> well, it should simply detect UTF-8 if that's your locale (when
> autotest doesn't interfere) but many users have other encodings
> already in use (KOI, EUC, ISO8859-x) and we need to be able to
> interoperate with them.

I think the best solution is to assume UTF-8, and use LC_CTYPE's
charset in case the filename is not valid UTF-8.

-- 
Robbe

pgpcrCZ8v835e.pgp
Description: PGP signature

_______________________________________________
Monotone-devel mailing list
Monotone-devel@xxxxxxxxxx
http://mail.nongnu.org/mailman/listinfo/monotone-devel