On Thu, Jun 18, 2015 at 3:11 PM, Michal Dziewonski
<
mdzie...@mozilla.com> wrote:
> I also have this:
>
>
https://support.mozilla.org/bn-IN/localization
Michal, what you provide in those two URLs constitute a dashboard. The
short version of what I see is that the group of individuals who
contribute to SuMo from India don't seem to have enough contributions
coming in for the bn-IN locale.
In other words, the bn-IN community has debilitated. The proposal I
see being discussed indicates that there is a general agreement (based
on data/study yet to be made available to a larger group) that this
lack of content in bn-IN can be easily replaced by
replacement/redirecting to content available in bn-BD.
This is an interesting scenario. It implies that the Mozilla community
and the staff members are putting in a precedent that content in two
locales are equivalent and replaceable based on unknown metrics. What
long term effect could this bring about?
Here's a small set of conjectures.
[+] bn-IN localization community fails to meet the string translation
release requirements for a specific release. Based on the precedent
set above, it is possible to consider a drop-down replacement and
shipping of bn-BD content for the application strings
[+] any other 'pair of locales which look similar' have this issue,
let's say ta and ta-LK, the precedent allows a similar set of
adjustments to be brought about
[+] with Mozilla setting a precedent as above, it would be a well
established guidance for other Free and Open Source Software projects
to consider likewise. For example, GNOME can now decide that it would
be appropriate to subsume bn-IN locale content with the bn one (if I
recall this correctly, Mozilla is the only upstream project with a
bn-BD and bn-IN. The pair is bn-IN and bn for everything else. One can
trivially look up the Mozilla bug related to the creation of these
locales for Mozilla and figure out why)
I am happy that you see the concern. Unfortunately, it would appear
that you are unable to differentiate between investing in a community
when it is sick and shipping a shiny product. The current set of
discussions way back in January didn't have 'experts' in locales or,
language technologies. If we really do want to seek them, then there
is a different and better way of doing this than sudden announcements
on lists.
In short, should this 'pilot' be considered and implemented, the
digital existence of bn-IN locale would be decimated. And this is not
hyperbole or, exaggeration of facts. It is how the long game will
eventually play out across the project communities.
I would be happy to seed content from en (en_US?) to bn-IN by means of
machine translation (<
https://wiki.mozilla.org/Intellego> ?) rather
than seeing short term gains make bn-IN a lost community.