Re: Translating schema files (was: Re: Survey results (yay!))

[Gil Forcada <gforcada gnome org>]

El ds 30 de 10 de 2010 a les 11:30 +0200, en/na Johannes Schmid va
escriure:
> Hi!
> 
> > So, just stripping out the gconf/dconf schemas we are getting 17% less
> > words to translate!
> 
> > If we do the same with the errors (much harder to do the analysis
> > though) we could nearly shrink the number of words to translate, at
> > least to 30% (so ~60k words less).
> > 
> > On the other side, or to further bold this argument, with the new
> > moduleset proposal made by the release team, more and more applications
> > are going to pop up, so more strings/words to translate...
> > 
> > Now that we have the numbers ... what do you think? Would it make sense
> > to propose the release team to ask to create different po files
> > depending on the string type (schema, error and general)?
> 
> Well, I disagree here: A complete translation of a desktop means to me
> that all user-visible strings are translated. And that includes error
> messages on the terminal as well as dconf schemas (that are shown at
> least in dconf-editor and possibly in the to be developed GNOME
> configuration tool).

Yes, a *complete* translation of a desktop involves translating
everything, *but* giving early feedback and actual results is priceless
if you are two translators and one of them has to struggle with git to
send the translations.

Hopefully with Damned-Lies with git this translator will have more time
to do their job instead of opening terminals :)

> The gtk+ properties are a bit different here because they are definitly
> only shown to developers and not to users (except in glade, but there
> user = developer).

I could also go further and separate the platform as a second-step
translation, that way a translator path for reaching the glorious 100%
should be:
- translate the ui
- translate the errors
- translate the schemas
- translate the platform

> Futher, I think extra po files complicate the build system quite a bit
> and I really don't want to do that for my modules (anjuta has about 20
> schema files though they aren't translated at all atm).

Developers will have to deal with it *once* and all other maintainers
will have to do the same, so you can cherry-pick their patches.

On the other side, this 35k words are a huge effort to translate, and
even if you skip them, having all modules 80% translated doesn't help
either. If you leave the schemas untranslated you may not notice that
anjuta developers have added a couple of strings since it will be
roughly 80% translated, and thus you will be missing some strings on the
UI.

> I feel this is the pseudo discussion for teams reaching yx % while it is
> far more important that it is convenient for the user. You can have 95%
> without translating nautilus and the user would still feel that half of
> the translation is missing. These numbers are just for our egos...

Yes, if you don't translate nautilus users will have a weird experience,
but if you make the big effort to translate 35k words and the users
don't see any translated string at all, they will close the session and
use another system.

It's not about egos, it's about splitting the work into small chunks so
that translators can start working on the stuff that 99% of the users
will see and then, if they have time (note that only 50 teams out of 120
reach the 80% mark) work on adding an extended experience with schemas
and errors translated.

> And 6% of the strings isn't really much. The difference between
> translating a word and a string is mostly marginal if the strings isn't
> five sentences long.

Yes 6% isn't that much, but the same can be said to the importance of
bugs, why have it then? Everything is a bug ... but if you mark a
crasher you expect that the developers fix this bug first before fixing
a minor.

Cheers,

> Regards,
> Johannes
> 

-- 
gil forcada

[ca] guifi.net - una xarxa lliure que no para de créixer
[en] guifi.net - a non-stopping free network
bloc: http://gil.badall.net
planet: http://planet.guifi.net