Re: Serbian (sr) language translation team: maintainer unresponsive

[Pablo Saratxaga <pablo mandrakesoft com>]

Kaixo!

On Mon, Apr 07, 2003 at 03:49:18PM +0200, Danilo Segan wrote:
> There are a few problems with the Serbian translation process for GNOME 
> 2 desktop & development environment.
>
> First is a completely absurd code assignment. First, code "sr" (ISO code 
> for Serbian language) is used for Latin transcription of default Serbian 
> cyrillic alphabet, and "sp" (this is not ISO code, and has nothing to do 
> with the language) is used for cyrillic (which ought to be default "sr").

Yes.

The reason of that is because XFree86 used those codes ("sr" for latin
script, and "sp" for cyrillic; used together with non assigned country
codes too) 

When some people contacted me to volunteer doing serbian translations
for mandrake tools there were already quite a lot of KDE translations
in latin (most of them in plain ascii, not even iso-8859-2, btw);
So, I've been faced with the dilemma of either using the right 
language code for cyrillic, or using "sp" and keeping "sr" for latin
script (as used by KDE).
Also, some people seemed to prefer usign latin on computers, others
prefering cyrillic... so, to keep everybody happy, the best was providing
both; and I requested translators to send me the po files in cyrillic
(as cyrillic->latin conversion is automatizable; the other way is not
("/dev/null" and similar things converted to cyrillic are not a wise thing)).
And so I used "sr" for latin and "sp" for cyrillic.

Note also, that at that time, utf-8 was not yet an option, so it was
impossible to have "sr@latin" and "sr@cyrillic" at a same time, the
need of different language codes was because of different, non
compatible, encodings (non compatible: impossible to convert in a losless
way from one to the other; while you can convert cp1251 <-> iso-8859-5
for serbian, you cannot between latin<->cyrillic (and the iconv //TRANSLIT
wasn't there yet, either)

Now things have changed quite a lot; and I wonder if latin serbian is
still as wanted as previoulsy (does KDE still uses latin?)
 
> So, the suggested resolution should be: "sr" for Serbian language (which 
> means also cyrillic alphabet), and "sh" for latin transcription.

That seems a good idea.
(if the translators agree, tell me when the switch is done, so I can do the
same change for mandrake translations, thanks)

> Tomislav, and will also ask for his approval to the job. Currently, 
> keeping up my translation current and useful is a real pain (CVS 
> updates, intltool-update and all the other stuff, which makes it hard 
> for many to have translated .desktop and .glade files ??? especially those 
> using binary distributions).

You can use the web status pages to get updated po files, and work on them;
it would be much less painful.
(in particular for translators that don't have CVS/intltool/etc,) 

> The best argument I can make for myself is the translation I've made 
> available at http://www.kvota.net/prevod/

I suggest you name the files not as xxxx.po but as xxxx-ll.po
that is particularly true in the case of Serbian as there would be
two files (one generated from the other, but two anyway):
xxxx-sp.po xxxx-sr.po or xxxxx-sr xxxxx-sh.po in case the language code
change goes on. 
 
> Cheers,
> Daneelo Shegun

Thanks

-- 
Ki ça vos våye bén,
Pablo Saratxaga

http://chanae.walon.org/pablo/		PGP Key available, key ID: 0xD9B85466
[you can write me in Walloon, Spanish, French, English, Italian or Portuguese]

PGP signature