[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-i18n-doc
Subject:    Re: Number localization in KDE4
From:       Chusslove Illich <caslav.ilic () gmx ! net>
Date:       2007-04-26 9:55:22
Message-ID: 200704261155.23039.caslav.ilic () gmx ! net
[Download RAW message or body]


> [: Nicolas Goutte :]
> The problem of being in the context is that the whole string (context
> and message) is handled at runtime. Also the msgctxt keyword is
> documented as  being small in Gettext, unlike in KDE3 where some
> contexts are very much  longer than the message to translate.

Although I'm not sure if you yourself are thinking of that, I'll just mention 
that any performance considerations due to the lookup is just superfluos when 
compared to other costs.

Also, just to make that clear too, from my point of view the usage of contexts 
in KDE3 is exactly what I want to keep.

> As for batch processing, I think that we two can do more batch
> processing by scripts (e.g. fuzzy in special entries for a very
> important need) than translators are able to batch process the
> other way round.

For centralized KDE translation, this is the real dilemma: have true alerts 
sometimes not raised (with comments), or have false alerts sometimes raised 
(with contexts). In the first case, the messages would have to be 
batch-fuzzed on comment change if deemed necessary, and in the second case 
batch-unfuzzed if the context change is irrelevant. Given the number and 
importance of this info to translators, I opt for the possibility false 
alerts.

-- 
Chusslove Illich (Часлав Илић)
Serbian KDE translation team

[Attachment #3 (application/pgp-signature)]

[prev in list] [next in list] [prev in thread] [next in thread]