Re: GNOME CVS (silent): bug-buddy pablo

[Pablo Saratxaga <pablo mandrakesoft com>]

Kaixo!

On Mon, May 21, 2001 at 05:38:43PM -0400, Jody Goldberg wrote:

> > >   po             : ChangeLog zh_CN.GB2312.po zh_TW.Big5.po
> > > Log message:
> > > fixed multibyte
> >
> > does this mean that we should be able to distcheck with gettext 0.10.37?
>
> It means that 0.10.37 needs work.
> This 'solution' feels like a kludge.  Stripping out 8bit characters
> to get some of the encodings to work seems very questionable.  Can
> we not add a lookup table to gettext to act as a backup to iconv
> failures translating accented characters ?

That problem is completly irrelevant and unrelated to the use
of gettext 0.10.37.
Note that I did only a few (2 or 3) such changes.
gettext 0.10.37 just make those cases more visible.

So, we have two different things to consider:
- use of gettext 0.10.37 or not (advantages: make big5 files readable,
  makes multibyte encoded files readable; disadvantages: big5 files made
  for 0.10.37 will fail for old gettext. However as big5 translators want
  to go for new gettext I think the discussion is somewhat over)
- inclusion of arbitrary non-ascii chars into text string the sources
  (so far, I've seen use of iso-8859-1 and iso-8859-2 chars, without telling
  the encoding in either case; but, from a strict philosophic point of vue,
  if those are accepted, why not euc-jp or koi8-r?)

Those two points are unrelated.

For the first point, I don't think it makes any real problem.

For thesecond point, just including an arbitray encoding without making
sure the program will be able to convert to the user's encoding is just
plain wrong (and was wrong before, when gettext didn't helped to find those
case); it is simply against the i18n idea.

> _______________________________________________
> gnome-i18n mailing list
> gnome-i18n gnome org
> http://mail.gnome.org/mailman/listinfo/gnome-i18n

-- 
Ki ça vos våye bén,
Pablo Saratxaga

http://www.srtxg.easynet.be/		PGP Key available, key ID: 0x8F0E4975