[GTK] Enable translations for WebKit2
Created attachment 191386 [details] Enable translation for WebKit2GTK+ Something that we might want to consider: custom rule that installs .mo files a second time but with a different name for WebKit2GTK+, which would need adjusting our usage of GETTEXT_PACKAGE as well, but would allow having different versions of wk1gtk and wk2gtk installed. I am not sure if it's really useful to have this flexibility - the Debian package sure would build both from the same source, so there would always be the same version of both in the archive, in addition, sharing is a good thing. But... something that needs considering.
Created attachment 191484 [details] Patch for EWS
Do you also need to update the paths listed in POTFILES.in? I think you can tell if it's working properly by running intltool-update -m from the directory containing POTFILES.in.
Unsure why the patch is not getting applied.
(In reply to comment #3) > Do you also need to update the paths listed in POTFILES.in? I think you can tell if it's working properly by running intltool-update -m from the directory containing POTFILES.in. Good catch! I haven't been able to properly test this locally because everything complains libc doesn't know the locale and it falls back to C, my system apps all work, though, I asked in #gnome-hackers but got no reply, let me know if any of you has a clue what may be causing this in my jhbuild env?
(In reply to comment #5) > Good catch! I haven't been able to properly test this locally because everything complains libc doesn't know the locale and it falls back to C, my system apps all work, though, I asked in #gnome-hackers but got no reply, let me know if any of you has a clue what may be causing this in my jhbuild env? Sorry. I'm pretty clueless about translations except for the basics about getting intltool working.
Created attachment 191501 [details] Patch for EWS
(In reply to comment #6) > Sorry. I'm pretty clueless about translations except for the basics about getting intltool working. Guess I'll make dist and try it out on a debian package ;)
I tested applying the patch locally with svn-apply and the error seems to be sr@latin.po and sr.po cause failed hunks, I'll see if I can figure out what that is.
Created attachment 191886 [details] Patch This has been properly tested, it turns out what was causing my "locale not supported" issues was setting LC_ALL explicitly in the wk2 process launcher. I believe that should not be needed, let me know if there's some use case that was supposed to support that I should test. I don't think the EWS will be able to apply this one as well, I haven't figured out why the sr*.po files cause trouble, I think we'll have to land this one without the help of the EWS and CQ... let me know if it looks OK to you
Comment on attachment 191886 [details] Patch Looks good, apart from the fact that the paths in the po files are incorrect. Maybe there's a way to automatically fix those. :(
Comment on attachment 191886 [details] Patch http://trac.webkit.org/changeset/145081