Re: New d-i translation scheme, who is a D-i i18n coordinator?
On Sun, Aug 29, 2004 at 10:08:15AM +0200, Christian Perrier wrote:
> However, I'm still convinced that the general coherency achived with the
> single file, acting like a big compendium file has still enough
> advantages for allowing us to think about better methods for handling
> its updates.
I'm with you, Christian :) I don't remember if I have already detailed my
vision of the Debian L10N infrastructure, first part of which is the step
you taken (a big po-file), but here it is...
<vision>
What I would like to have ideally, would be putting the po-files into a
database (parsed) with references to which templates they belong to. Each
string gets it's own translations, one of which is set to "correct".
Using this database it should be rather easy to generate
1) big po-file for each language
2) po-files for each available template and language.
Add some kind of interrogation between the database and CVS (for those
who'd like to use it) and an XML-interface for the package maintainers,
which would allow to submit templates and templates' updates to the system
and retrieve the generated po-files.
Main advantage: we'd keep track of all the changes and would also give
credits everythere they are due. We could avoid using msgmerge and give
the possibility to work using a web-browser i.e. not necessarily needing a
KBabel or a text editor.
</vision>
I just have to get to programm this ;) You've done the first big step, I
think it's the right direction.
--
Nikolai Prokoschenko
nikolai@prokoschenko.de / Jabber: pronik@jabber.org
Reply to: