|
|||||||||||
|
[RFC] Marking D-I strings for translation priority
From: Christian Perrier <bubulle(at)debian.org>
Date: Sat Nov 10 2007 - 01:46:53 EST
From an earlier discussion in -boot, mostly with Frans Pop, the following idea came up: > Quoting Frans Pop (elendil@planet.nl): > > > > And the splitting out into sublevels could be done by again first That seems feasible as long as each string in PO files is "marked" with a comment. The point is to categorize strings in five categories:
Strings could be "marked" in such way: Template: debian-installer/localechooser/title Type: text # This menu entry may be translated. # However, translators are required to keep "Choose language" # as an alternative separated by the "/" character # Example (french): Choisir la langue/Choose language # sublevel1 _Description: Choose language .../...
Template: debian-installer/locale
That would give in PO files: #. Type: text #. Description #. This menu entry may be translated. #. However, translators are required to keep "Choose language" #. as an alternative separated by the "/" character #. Example (french): Choisir la langue/Choose language #. sublevel1 #: ../localechooser.templates-in:1001 msgid "Choose language" msgstr "Choisir la langue/Choose language" .../... #. Type: select #. Description #. sublevel2 #: ../localechooser.templates-in:13001 msgid "Choose a locale:" msgstr "Paramètres régionaux :" The "marker" is yet to be defined. Something as simple as possible with the guarantee to not appear in "regular" comments in enough. Please note that this *will* increase the size of templates file AND PO files but NOT the size of udeb packages as the templates files in them is generated without comments. So, the size of D-I *source* packages will be increased but not the size of binary packages. Once all strings are marked, hacking would be needed (mostly with msggrep) to split out the master PO file into 5 sublevel files containing the strings for each sublevel. An late rhacking will be needed to l10n-sync so that strings in packages can be updated from these sublevel files. That would allow new translators to focus their work on the most "important" strings first, then gradually work on more "obscure" strings. As a consequence, we (D-I team) could decide of better criteria to activate languages (this was the point of most of the discussion I had in list with Frans). Unless someone actively disagrees, I will start working on marking our 1686 strings as soon as possible (that's quite a huge work), so plaese comment quickly, particularly if you dislike the suggestion. -- To UNSUBSCRIBE, email to debian-boot-REQUEST@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.orgReceived on Sat Nov 10 01:47:08 2007 This archive was generated by hypermail 2.1.8 : Wed Mar 19 2008 - 03:19:07 EDT |
||||||||||
|
|||||||||||