Quoting Denis Barbier ([EMAIL PROTECTED]): > On Sun, Aug 29, 2004 at 08:39:54AM +0200, Christian Perrier wrote: > [...] > > The "single file" system for translations is currently mostly used by > > translators who work alone (except French, which is currently often > > used as a test lab for new work methods for teams). > > This "single file" system can hardly be seen as a work methods for > teams, there is a single responsable for the PO file and proofreaders' > work is much harder.
Well, just like big po-debconf files....which has been proven to work, given the number of xfree86 debconf translations.... For sure, D-I translation is one order of magnitude above this with about 1200 strings. 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. In my opinion, with such file, the whole team is responsible for the translation of course. The team just needs to establish a protocol for updates and review....which is not very different from the mathod we already have for other files. Moreover, there are enough identical or nearly identical strings in D-I and thus updates such as the recent templates Sven added to nobootloader automagically inherit similar translations from other packages. > This is the wrong way, large files are usually split into small pieces > in order to be managed by several persons (like the installation > manual). > Most drawbacks against the old scheme were already fixed by > http://lists.debian.org/debian-boot/2004/03/msg05834.html Well, the current status of joey's PO repository is not easy to use....It is very slow for updates as it works as an intermediate between the user and the "real" SVN repository. And, as Joey mentions, it requires maintenance when new packages are added. At the moment, the single file model is not mandatory as it is completely compatible with the multiple files model. As already mentioned, the work on French updates will be used as a validation lab for problems induced by this model.....which may benefit other teams. Norwegian teams already work with only one file for D-I for months (they used Petter's gettext-helper)...It does not seem to have induced that much problems. The Greek translators also seem happy with the new scheme..... Remember that, now, D-I translations do not change that much. Nearly all recent changes were most often quite trivial ad did not often require deep review. We still have time for more deep thinking when it comes at team's work. However, I strongly want to simplify things as much as possible for NEW translators, most of them not coming from the Debian world. This is the main purpose of this proposed (and somewhat pushed) scheme. Another goal is a simplification to the current "stages" concept which needs a lot of explanation each time a new language comes into the game. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]