Re: .jp Localization of apps/libraries. (was "Re: [Gnome-print] Re: [ynakai@redhat.com: Re: Gnumeric bug 15607]")

[Owen Taylor <otaylor redhat com>]

Miguel de Icaza <miguel@helixcode.com> writes:

> > But currently, we aren't there yet, and won't be there for the
> > GNOME-1.4 timescale. We need to make sure that Gnumeric (etc.) does
> > print in Japanese, and since we don't have the complete tools for
> > printing in Unicode, that means adding EUC-JP support.
> 
> Yukihiro's patch did not add any printing facilities, it just fixed
> the display output in a way that was pretty hard to understand.  No
> ChangeLogs, no description of the changes.
> 
> That makes our life as maintainers very hard.  If Yukihiro wants to
> keep a patch to do those things in his local copy, that is fine with
> me, but it wont go into Gnumeric, as we are trying to move gnumeric
> and gnome-print into supporting utf-8.  

Since I was brought into this thread fairly late, I don't know the
exact changes that started the discussion. However, it is certainly
true that most Japanese computing still works with EUC-JP. The fonts
available, the files people have, the locale that they run in, the
encoding that printers expect...  So, in order to work with Japanese
systems, compatibility with EUC-JP is a requirement. 

Just being able to display a few Japanese characters via Unicode is
not sufficient.
 
> I have a bit of experience from the patches from some of the hackers
> in the Japan office of Red Hat: they sent a set of patches to Radek
> for GtkHTML that were completely broken, again, no ChangeLogs, and
> barely any testing was done on them.  When we decided to revert them,
> the contributor was very upset.  

I firmly believe that it is the responsibility of the maintainer 
to make sure that any code going in is up to standard.

If an incoming patch doesn't meet the requirements, then you
should get back to the patch submitter and say what needs to
be improved, not apply the patch, and then later revert it because
you decided it isn't any good.

If you don't have sufficient experience with the traditional Unix
locale model to evaluate a patch, feel free to send it for me to
comment. Hopefully I can help come up with a solution that will work
as we transition from

 a) the current system of locale-dependent encodings
 b) the future system of Unicode / UTF-8 everywhere
 
> One of Yukihiro's previous mails said he was not working on this
> during his work hours, and hence had no time to submit a ChangeLog
> entry.  This is not the kind of code I would like to maintain, and
> certainly the lack of testing is disturbing.
> 
> I hope Lauris will keep working on improving our utf-8 support, and
> help migrate Gnumeric together with Jody to a bring utf-8 world
> (should not be too hard), and this is of course, fully independent of
> the GNOME 1.4 release schedule.

Its good to get experience with UTF-8, and I'm glad to see people
working on it. That being said, I think that until gnome-print is
using Pango, and other places are using the final GLib unicode
interfaces, not the obsolete libunicode, there is quite a bit of work
remaining to be done.

Full Unicode support for GNOME has to be thought of in a GNOME-2.0
timescale - the facilities just aren't there in what we have now, or
will have until we move to the GNOME 2.0 platform.

Regards,
                                        Owen