[Po4a-devel]Re: [Translate-devel] Presentation of the po4a project

[Denis Barbier]

On Fri, Nov 12, 2004 at 04:08:02PM +0200, Dwayne Bailey wrote:
> On Fri, 2004-11-12 at 12:01, Martin Quinson wrote:
> > At the first glance, I'd say that there is a bunch of functionalities in
> > common in both projects. The main differences are the targeted formats. For
> > now, po4a has working modules for sgml (both DebianDoc and Docbook DTD), XML
> > (a bunch of DTDs), groff (for man pages), pod (for the perl documentation),
> > and the documentation strings of the 2.4.x kernels. Another difference is
> > that po4a is in perl while Translate is in python. :)
> 
> Well it seems that together we are covering more bases with this PO
> format thing :).  Between the two projects we've got most localisable
> stuff covered.

Hi there, I am another po4a guy, but this mail is about something
different.  In Debian, no locales are shipped and users have to select
which ones to install.  But as you know, adding new locales may be
troublesome because GNU libc may need to be recompiled (more precisely
the localedef executable), which is no fun (at least for Debian guys,
Gentoo addicts may have a different opinion ;)).

So I hacked glibc to extract the bits necessary to compile localedef,
and while I was on it, also packaged new locales data files by including
many bug reports against upstream and Debian bugzillas.  Sources and
binary (only i386 at the moment) packages are available from
  http://people.debian.org/~barbier/belocs/pool/main/b/
These are unofficial packages and I do not intend to push them into
Debian for the moment.

Compiling these packages on non-Debian systems may be tricky, do not
hesitate to ask for help if you are willing to do so.
All the *_ZA locales you sent to glibc bugzilla are included, but you
forgot to specify which encoding should be used, so here is what I
added to localedata/SUPPORTED:
  af_ZA.UTF-8/UTF-8
  nso_ZA.UTF-8/UTF-8 nso_ZA/CP1252
  ss_ZA.UTF-8/UTF-8  ss_ZA/ISO-8859-1
  tn_ZA.UTF-8/UTF-8  tn_ZA/ISO-8859-1
  ts_ZA.UTF-8/UTF-8  ts_ZA/ISO-8859-1
  ve_ZA/UTF-8
Some values are arbitrary, I tried to follow GNU libc naming conventions
and anticipate what GNU libc developers will decide, but in any case you
should certainly add a '% Charset:' comment in your locale files to list
valid encodings.  E.g. I am not sure if CP1252 is the best choice for nso_ZA.

Denis