Le Tue, Jul 02, 2002, à 11:13:59AM +0100, Alan Horkan a écrit:
> > While doing the French translation today, I noticed that the fuzzy
> > translations can overinflate the score (while puzzling end-users). I'll see
> > how I can update the graphics scripts (and CTR) to better take this into
> > account.
>
> it also looks really weird that en_GB is only 1% but as the person at
> Abiword who started en-IE, en-GB, and then saw en-AU, en-NZ and one or two
> others i understand the score (and developed a huge distaste for
> localisation, it is too difficult to maintain the same tiny change across
> 5 locales, while also trying to translate Abiword into a minority langauge
> like Irish)
en_XX (where XX != US) is indeed a special case.
> AFAIK gettext does not do locales very well
> recently one of the abiword developers was adding all sorts of
> unusual and esoteric langauge codes not just of the form xx-XX
I don't know whether gettext is better or worse than alternatives... and how
to improve it. I think it does a fair effort and achieves a not too bad
result (though the fuzzy merges sometimes leave me perplex).
> you would be amazed the huge variety of people out there who would like
> abiword to work in their language (although you probably encountered about
> half of them looking for Dia in their langauge.
never heard about them... though I notice we ship quite a few .po files.
-- Cyrille
--
Grumpf.