migrating 28-30 locales from mozilla-central

[John O'Duinn]

hi;


If you are involved in any of the following locales, please read on.
Otherwise, you can hit delete now!


af
ak
as
br
bn-IN
da
eu
gd
gu-IN
hy-AM
is
ka
kk
km
kn
lg
mai
mk
mn
mr
nso
oc
or
rm
si
son
sw
te
ta-LK
zu


In bug#657789, l10n-drivers gave RelEng this list of locales which had
already stopped developing on mozilla-central, and were safe to migrate.
Because its been a while, I am posting here to remind everyone, and
avoid anyone being surprised when this happens in the coming days.

For each of these locales, we plan to stop generating nightlies on
mozilla-central, and will generate custom updates to migrate any users
on that mozilla-central locale to a mozilla-aurora build for that
locale. (By request, af and oc will migrate to en-US and remain on
mozilla-central, see bug#657789 for details).



If you are involved in any of these locales, you had previous signed off
on this, and you now have any questions/concerns, please ask in
bug#657789 or email me directly.

Thanks
John.
=====
Background: As part of Mozilla's conversion to rapid-release-model, most
localization work switched focus away from mozilla-central onto
mozilla-aurora and mozilla-beta. A handful of locales need to stay on
mozilla-central, to help engineering and localization-edge-cases, but
all other locales will be moving from mozilla-central to mozilla-aurora
as soon as they are ready.

Mozilla generates nightly builds, and updates, for each locale on
mozilla-central and mozilla-aurora as well as for the beta locales on
mozilla-beta. Trimming any unwanted localized nightlies from
mozilla-central is a great help to our nightly compute load.

Trimming unwanted localized nightlies also reduces confusion for
localizers, and reduces the number of string landings a localizer has to
do, because they will no longer need to cross-land onto mozilla-central
as well as mozilla-aurora.

=====

[Axel Hecht]

This data is old, and whatever we thought back then isn't valid today.

I've been collecting some data to inform our proposal here:

How many weeks has a locale not pushed to l10n-central.
How many users do we have on the nightly channel.

I.e., where do we have interest in the wild, and where do we have
interest in the l10n community.

I'm proposing to cut at: More than 22 weeks of no pushes today, and less
than 10 users. That's way on the safe side, and continues locales where
we see enough users on nightly to use that for community building.

That would leave us with 37+en-US locales on Nightly, cutting 51. 37
locales includes our top locales, and is a pretty healthy set.

What we don't have at this point is testing of RTL, neither on the user
nor the contributor side. That's a bit concerning. So I'm moving the 3
RTL locales in the bucket to keep. Keeping 40, dropping 48.

If you have comments, please share them 'til mid-next week, so that we
can move forward.

Locales to build:

ar
ast
cs
de
en-GB
eo
es-AR
es-CL
es-ES
es-MX
fa
fr
fy-NL
gl
he
hu
id
it
ja
ja-JP-mac
kk
ko
lt
lv
nb-NO
nl
nn-NO
oc
pl
pt-BR
pt-PT
ru
sk
sv-SE
th
tr
uk
vi
zh-CN
zh-TW

Locales to stop building:

af
ak
as
be
bg
bn-BD
bn-IN
br
bs
ca
cy
da
el
en-ZA
et
eu
fi
ga-IE
gd
gu-IN
hi-IN
hr
hy-AM
is
ka
km
kn
ku
lg
mai
mk
ml
mn
mr
nso
or
pa-IN
rm
ro
si
sl
son
sq
sr
ta
ta-LK
te
zu

[chris hofmann]

Last call for input. releng will be moving forward on shutting down the
"Locales to stop building" that Axel lists below unless there are
reasons not to posted here in the next few hours.

-chofmann

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