hi Andrew
> Yeah, indeed. Though gettext does do some fuzzy matching on things, so
> they might not all be completely lost
Yeah - I saw a lot of those "fuzzy" entries in the German translation file.
These were mostly the parts where I "retranslated" the texts :-)
> I'd be happier to make changes like this when I'm not making
> functionality changes at the same time. So maybe after 0.9.8 is
> released I can start accepting patches like this.
Well, you'll need a lot people of different countries to translate the stuff
after accepting such changes!
You also need them when you release a new version which changes or adds a lot
of strings - so I thought it'd be better to ask them once for a new
translation instead of twice...
But it's you're decission ,-)
> For now, perhaps read
> up some further on the capabilities of gettext in this regard:
>
> http://www.gnu.org/software/hello/manual/gettext/Plural-forms.html
I read this - and understand it, in principle.
But what I not understand is how the message strings in the PO files should
look like for languages with "strange" plural rules.
There are always (only) two msgids given to ngettext, so how does it create
msgid's / message stringnames for the other cases (e.g. when you need three
texts).
Does it automagically add a "1", "2", ... to the original msgid and create a
new one?
Or does it add that number to the msgstr?
In any case the usage of that extended gettext functionality makes it more
complicated for the developers...
kind regards,
Matthias
> Yeah, indeed. Though gettext does do some fuzzy matching on things, so
> they might not all be completely lost
Yeah - I saw a lot of those "fuzzy" entries in the German translation file.
These were mostly the parts where I "retranslated" the texts :-)
> I'd be happier to make changes like this when I'm not making
> functionality changes at the same time. So maybe after 0.9.8 is
> released I can start accepting patches like this.
Well, you'll need a lot people of different countries to translate the stuff
after accepting such changes!
You also need them when you release a new version which changes or adds a lot
of strings - so I thought it'd be better to ask them once for a new
translation instead of twice...
But it's you're decission ,-)
> For now, perhaps read
> up some further on the capabilities of gettext in this regard:
>
> http://www.gnu.org/software/hello/manual/gettext/Plural-forms.html
I read this - and understand it, in principle.
But what I not understand is how the message strings in the PO files should
look like for languages with "strange" plural rules.
There are always (only) two msgids given to ngettext, so how does it create
msgid's / message stringnames for the other cases (e.g. when you need three
texts).
Does it automagically add a "1", "2", ... to the original msgid and create a
new one?
Or does it add that number to the msgstr?
In any case the usage of that extended gettext functionality makes it more
complicated for the developers...
kind regards,
Matthias