Quoting Thomas Diehl : > Hi, > > I'm attaching a mail by Andreas Pour on how he thinks > we could improve the GUI translation process. I will also > forward the follow-up mails that came of this so far. > > Please cc all replies to Andreas since he is probably not > subscribed to this list. > > Regards, > > Thomas > > ==================BEGIN FORWARDED MESSAGE================== > >Date: Sat, 18 Nov 2000 12:57:29 -0500 > >From: Andreas Pour > >To: Thomas Diehl > >Subject: Tranlsation Website > > Hi, Thomas, > > I have an idea for improving the translation process and getting more > people involved. I outline the idea here, I hate to be brief but I > don't have very much time today but wanted to get this out there: > > (1) script updates daily from CVS and inserts all "original" message > strings and their translations in a database (HTTP-accessible); > (2) anyone can look up missing translations in the db and enter > translations; > (3) only kde translation team members can "approve" a translation; > (4) at the end of each day .po files are automatically generated from > the database and changes are added to CVS. I think (4) is not so good. I don't want a script that break my translation because there are not enought context info. I cannot, every time I "approve" a translation, know if there "special" different translation somewhere in the CVS. I think the creation of the database is a good thing and the possibility for nomral user to ask to change (or propose) some translations is good too. I think we can easly add to next generation kbabel (i.e. kbabel with searchengine plugins) an engine the browse the "internet" database (best option is to download it once a day to speed up the database search). Scripts are good but they MUST NOT modify a non fuzzy or untranslated string!! Bye Andrea Rizzi