Re: GCC2 native language support now available

[Paul Eggert <eggert at twinsun dot com>]

   Date: Sun, 27 Sep 1998 13:59:12 -0400
   From: =?ISO-8859-1?Q?Fran=E7ois_Pinard?= <pinard@iro.umontreal.ca>

   Let me suggest that you use `cc', with no capitals and no prefix `g',
   for a single domain name.

I like this suggestion, but it has some technical consequences: if
gcc2 and egcs both use `cc', and if both install translation files in
the usual places, then gcc2 and egcs would share translation files.

If done naively, this would cause problems, since installing egcs
would overwrite gcc2's translation files, which would cause a
French-locale gcc2 to sometimes output English instead of French.

One solution is to share the `cc' textdomain between gcc2 and egcs.
This approach will simplify the translators' job, since it will mean
they have just one .pot file to translate instead of two similar ones.
However, it will mean that gcc2 and egcs will need to coordinate that
textdomain.

I hope that gcc2 and egcs development will merge soon; this will
simplify matters greatly.

   The distribution you made available apparently does not contain a POT file.

That's because GCC alpha snapshots don't contain files that can be built
by maintainers, and the .pot file is built automatically from the sources.

For your convenience I placed a compressed copy of that file at:
ftp://alpha.gnu.org/eggert/gcc.pot.gz

   we are not set for handling dates as versions, or to compare
   dates to version numbers and decide of the chronological order.

Sorry, I don't easily see how to shoehorn GCC's version numbers into
the forms currently allowed by the translation project.

How about if we set PACKAGE=cc and VERSION=2.8.1.19980813 for the
August 13 snapshot?  This numbering scheme sorts lexicographically in
the proper way.