Right set of locales to build on central?

[Axel Hecht]

Hi all,

I wanted to open a discussion on what's the right set of locales to
build on l10n-central for Firefox.

Many teams don't actively work on there, so spending computing and disk
resources on that might not be utterly worthwhile.

On the other hand there's a high value in those localization teams that
actually keep up with central and make sure the code that's there is
properly localizable.

We can probably discuss that independently for dashboard and for
generating actual builds, with nightly updates et al.

What do you guys think?

Axel

[Ehsan Akhgari]

Up until the release of Firefox 3.5, I used to translate both 1.9.1 and m-c,
and ported the work between these two branches for Persian (fa), so they
were pretty much in sink. I haven't been able to do much l10n work after
3.5, but I'm planning to keep both 1.9.2 and m-c localizations updated, so I
appreciate if you can include fa builds from l10n-central.

I think maybe we should make this thread an "opt-in" thread so that we can
pick up only those locales which actually work on l10n-central.

--
Ehsan
<http://ehsanakhgari.org/>

> _______________________________________________
> dev-l10n mailing list
> dev-...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-l10n
>

[wladow]

If you want to filter out several locales, my initial thought:

SELECT locales WHERE missing_strings < 100 AND latest_push > -30d

where the condition values may emerge from this discussion. By settings
a "good" values you can filter out locales not being actively worked on
along with keeping those that don't need special attention from a
localizer in terms of untranslated strings even if there wasn't any push
in the past few days/weeks.

Not sure where to get those data (the dashboard?) and how/whether is it
possibile to incorporate this into our build system, though.

--
wladow

[Дамјан Георгиевски]

> I wanted to open a discussion on what's the right set of locales to
> build on l10n-central for Firefox.
>
> Many teams don't actively work on there, so spending computing and
> disk resources on that might not be utterly worthwhile.

The main problem might be that different mozilla projects might work
from different repos. For example, we only worked on Firefox thus far,
so the release-1.9.x has been where our attention has always been.

But if we worked on other projects... well, for some of them it wasn't
known which repo they will be created from... so ... while curently it's
a pain for me to work with both the release-x.x and l10n-central trees,
if we get some commitment from people wishing to help (not likely) with
some other projects, l10n-central might get interesting again.

> On the other hand there's a high value in those localization teams
> that actually keep up with central and make sure the code that's there
> is properly localizable.

--
дамјан ( http://softver.org.mk/damjan/ )

Hi! I'm a .signature virus! copy me into your .signature file to help me
spread!