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

List:       kde-i18n-doc
Subject:    Re: [Idea] Translation manager
From:       Nicolas Goutte <nicolasg () snafu ! de>
Date:       2004-09-08 9:32:36
Message-ID: 200409081132.36908.nicolasg () snafu ! de
[Download RAW message or body]

On Wednesday 08 September 2004 10:51, Rinse de Vries wrote:
> Op woensdag 8 september 2004 09:06, schreef Nicolas Goutte:
> > That would lead again to runs to 100%, like it appears to already exist.
> > That is not in favor of translations.
>
> Why?
> I think it can release the presure of translators to get a 100% translation
> for a major release, because they are aware that their users can update
> their translation any time.
> If a team does not succeed in translating a certain application 100% (like
> kommander, that created about 400 new strings close to the deadline of kde
> 3.3), there is no need to rush. Just inform the users about this, and
> finish the translation in your own time. When it is finished and checked,
> users of the application can update easily.

But isn't it exactly with partial translation that you need to be able to 
update?

If a user has a 100% translation (or even 99%), he will not much care to 
update. But if it has only 25% and he knows that the current work is about 
40%, he might be more interested.

>
> > (And I thought that this ne
> > wmehcanism was to help with translation. If the translation is already
> > good and 100%, there is not much more need of such a mechanism.)
>

> No translation is free of errors, and most translation teams don't release
> a 100% translation in all modules.

That is why it should not be temptative to reach 100% with low quality 
translations.

> Everytime a kde-release gets out, in the last weeks there always are a lot
> of mails from teams asking to delay, flames about the continious flow of
> new strings and fuzzies, etc..

Sorry, I do not think that it would stop, not only because it is the full 
releases that get packaged and shipped by distros.

> With an easy mechanism to update translations, teams that can't reach 100%
> for a certain application/module don't have to rush things, they can
> release the complete translation when they feel fit.

A useful argument: it could be downloaded even after the last release of a 
stable branch.

>
> There is also another benefit:
> At the moment all translations all modules are included in 1 package.
> That means, if e.g. kdepim or kstars has a intermediate release, you can't
> upgrade your translation, because there is no intermediate i18n-package
> available. That means that users that install a newer version of e.g.
> kstars get a funny mix of English/Dutch, which is not very useable.
>
> The translation teams that provide an updated kdepim/kstars can publish
> that, and users can upgrade the translation of thos specific application
> without having to compile anything, or risking to break other translations
> while upgrading *all translations* to get an up to date kdepim/kstars

I suppose that whoever develops this application, he/they will really have to 
design how to do version matching. Otherwise I fear that it will become a 
nightmare very soon. (CVS HEAD, stable branch, special releases...)

(By the way: Kate has already a download mechanism for highlight description 
files. So perhaps it is code/design to start from.)

>
> Rinse

Have a nice day!

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