Hi, At the moment most of the content in translation commits is related to file position for the translatable strings.
As an example: http://git.gnome.org/browse/seahorse/commit/?id=66f3c65fd7ef596062b7c7f1f127ca70e95c6a16 -#: ../daemon/seahorse-daemon.c:69 +#: ../daemon/seahorse-daemon.c:64 msgid "Do not run seahorse-daemon as a daemon" msgstr "Nu executa seahorse-daemon ca serviciu" -#: ../daemon/seahorse-daemon.c:99 +#: ../daemon/seahorse-daemon.c:94 msgid "couldn't fork process" msgstr "nu s-a putut crea procesul" -#: ../daemon/seahorse-daemon.c:105 +#: ../daemon/seahorse-daemon.c:100 msgid "couldn't create new process group" msgstr "nu s-a putut crea un nou grup de procese" Endless scrolls of text with no value. The commit I'm linking to just adds a translation to one single string. Try and find it in there :) There are several downsides to commiting these numbers: 1) they make reviewing changes hard (both on l10n.gnome.org and by running "git diff" locally) 2) they increase the size of the repositories with useless data that will be carried around forever 3) they numbers are generally valid only for that specific commit. Once a new line is added to the source .c/.py/.etc. they become mere position indications. All text editors permit text search, and searching for the string itself will likely bring more accurate results than jumping to the file position. I'd like to know what is your opinion on using a git hook/other mechanism to filter these line numbers. -- . ..: Lucian _______________________________________________ gnome-i18n mailing list gnome-i18n@gnome.org http://mail.gnome.org/mailman/listinfo/gnome-i18n
