Gasper Zejn wrote: > Dne sreda 28 februar 2007 10:16 ste napisali: > >> As a ex-developer: >> I think it's important consider the implications for how you make >> projects run. A major positive feature at the moment for many projects >> is that you can simply check out your code from version control and run >> Pootle on top of that. Rearranging directory structures would make it >> harder to commit back. >> > > True, it would make harder for Pootle to commit back, but by looking into > subversion properties, a script could merge translations with checked out > sources. This would solve the risk of allowing Pootle to commit, since Pootle > would not commit anymore. The localization coordinator would, since it's the > coordinator's job to know what he committed. This is where Pootle needs to be > very careful, otherwise it could be greeted with same disrespect as Rosetta > once was. > > Therefore I would change "harder to commit back" to "require translation > coordinator to commit back". And this is a good thing, since upstream may at > any time ask translator about a specific commit, and he needs to know what he > is being asked about. > Aah, but can the translation coordinator still commit back through the web interface? I think its good to separate out the issues - rights as to who can commit back can be done through a rights system, and disabled for projects who don't want it. Requiring someone to commit manually is fine, but I think its good if that person can do it through the Pootle interface. I agree that caution is required here, but the caution needed is primarily on the part of those people running Pootle servers, making sure they are set up properly. The Pootle code just needs to enable what people need to control. >> I think the idea of using file properties with filesystem attributes or >> subversion attributes is great >> Nice to see your work progressing - I wonder if some of your code >> rearranging could be folded back into the standard Pootle branch to >> prevent the Django branch diverging from current development >> > > At first I tried to change code as less as possible, but then I figured out > since I'm already migrating, I would have to adapt code so much it didn't > really matter if I reimplement it. That's because Pootle had the 'project.py' > file, where everything was concentrated in TranslationProject object, and > Permissions and Goals need to be detached from TranslationProject. > Yes, an ugly mess I admit :-) > There are parts that could be ported, namely some lazy-fied methods of > TranslationProject, eg. for loading stats. They are usually mentioned in > commit messages. But that's a very very little part of all the code. > Well I think your restructuring might still be helpful, but that's up to you and Friedel etc to decide...
Thanks for all your work David ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Translate-pootle mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/translate-pootle
