I sadly won't be able to assist to this conf, and no member of the french translation team will be there, IIRC. Sad thing.
You may want to check w.d.o/french/intl/french That's in french only, but I guess you should get the essence as latin-based native speaker. If not, I may consider translating this even if it's somehow very specific to our team. What I want you to get from this page is the organization we cam with to handle efficiently the translations in debian. That's based on a loosely split of the efforts in several sub-project (web, ddtp, po-debconf), and at the same time in several efforts (i18n, even if our problem are far more simple than asian people, translation, review). All this done with classical mail title, so that each people can do exactly what he wants in the project without spliting the ML. My personal opinion is that the french team is not anymore in an i18n process, thanks to our predecessor work, and becomes not even in a pure l10n process. Getting new translation is not my [main] goal anymore. I want to make sure that the stuff already translated is correct (review, very important since the written french language is hard even for native speakers) and uptodate (need of specific tools to detect when a translation gets outofdate, warn the assigned translator, and ask for an NMU to the other if it's not enough ;works for webpages, still to do for other materials). So, logistic issues are critical to me, stressing the need of a translation tag in the BTS or other kind of support from the PTS. But this is not all. Our team is also very concerned by making stuff translatable (I know, only a part of i18n). po-debconf was made by Denis, coordinator of our team, and some of us dedicace quite a large amount of time to let the packages using debconf moving to this system (along with portuguese people IIRC). You may also say a word about l10n-check, a script developped in intern to catch to most common errors in the common format. po4a is also worst mentionning I guess (quite logical, that's my baby :). Its meant to ease the maintainance of documentation translation by using po files as main translation container. That way, when the original gets modified, the update is veeeery eased. And getting stats about translations is also a new feature. This project may not be mature enough for a wide use, but is definitivly promising in my mind. Ho, and we do have a shared dictionnary under w.d.o/intl/french/dico. That's a simple set of wml macro allowing to make a kind of encyclopedia. English words are explained in details, and some possible translation are proposed, with the arguments pro and against each possibility. This is quite new and not that much words are documented, but it's growing quite quick. In future work, you can also mention i18ning the News.Debian.gz, since it's probable that in a near future, most debconf notes will be converted to that system. i18ning deb-log is also a chalenging issue, when it ever gets written. I think that po4a may help, but it still has to be proven. That's it, dude. Have a nice trip. Bye, Mt. On Tue, Jul 08, 2003 at 06:44:01PM +0200, Javier Fernández-Sanguino Peña wrote: > Hi there, > > I'm going to give a talk at debconf regarding i18n/l10n projects "early" in > the morning sunday [1]. I wanted to know if people at -i18n feel that the > following issues which I wanted to talk about cover sufficiently l10n/i18n > in Debian (sorry if it's not fully understandable, it's a draft atm): > > - Intro: why is i18n/l10n important for Debian? > > - l10n/i18n projects in Debian: > - Webserver content translation > - l10n/i18n for Debian-specific programs (dpkg, apt, d-i, dselect...) > - po-debconf translation > - Debian package descriptions (ddtp) > - Debian-specific manpages > > - Generic issues: > - translator status (not a DD?) > - multiple interfaces for similar tasks (CVS, BTS...) make > it very difficult to contribute. > - Problems in CVS access (proxies) > - Problems in BTS use (translation tag?) (sample: dpkg has > 14 translation-related bugs, apt 5, how do I see the bugs > that affect a given language?) > - Which are the "officially supported" languages? > - Even if things are translated, translations are not always > distributed properly, usefullness drops (#64278, #172482) > - UTF-8 support? > > - Missing pieces for a fully i18n/l10n project (and OS): > - Single clearing-house for translations: simplify i18n work. > (look #105337) > - Provide tools for translation teams to use, like Debian-specific > translation catalogs/dictionaries. > - Full support of l10n in core system > (that means 'base', but one could include also Desktop-aware stuff > such as 'menu', it also includes system messages like init scripts) > - Easy setup of language environment for the user (just once, > on installation, a user selects a language and that setups > the task as well as the default locale) > - I18n BTS: allow translation teams to handle workflow through it > > - Some proposals: > - when creating the distribution provide translation bugs to > be fixed in freezes > - translators should be given DD status, but with different > privileges (no uploads but CVS commits) (Alioth?) > - Add a generic translation/'language' tag for the BTS > (so that translation teams can check bugs related to their > languages) > > ¿Ideas? ¿Opinions? > > Regards > > Javi > > > [1] More precisely at 10:00-10:45. Since the saturday dinner will be > "special" I don't know if many people will show up then. Or, if they do, > they will be able to pay attention :-) -- If the automobile had followed the same development cycle as the computer, a Rolls-Royce today would cost $100, get a million miles to the gallon, and explode once every few weeks, killing everyone inside.
