On 2012/02/10 06:50, Rob Weir said:
On Thu, Feb 9, 2012 at 5:28 PM, Michael Bauer <fi...@akerbeltz.org>
wrote:
Putting it the other way, would you give just anyone the right to
commit new (structural) code?
Community translation has its good sides and it's bad sides. What
will undoubtedly happen is that there will be disagreements over
style, terminology and orthography. Unless you have the option of
making someone the "arbiter" of that locale, you end up with the most
nauseating hotchpotch of style, terminology and orthography. Some of
the Google interface languages are like that. I have read a little in
the mailing list. Openoffice uses a pootle server for translation[1].
It might be worth to think if that might be useful for us too.
Apache projects tend to look at this slightly differently.
Traditional, vertical organization: Someone (or group of persons) is
in charge and decides what will be done, and then delegates that
authority to other individuals who have that authority and decide such
matters. Parts of Apache are like this, at the foundation-level. For
legal reasons there are individual officers with formal
responsibilities and authority that supports those responsibilities.
But within a project, it is more like horizontal organization,
consensus based. The community agrees that it wants consistency in
terminology in the UI translations. Someone volunteers to define a
basic terminology, the community adopts it as a "coding standard" and
agrees to treat deviations like a bug. Someone might volunteer to
review these items and fix errors and general facilitate consistency
in the translations.
So I think you end up in the same place in both models. You achieve
consistency. In both cases there is leadership, but in the 2nd case
there is no appointed leader with authority. There is just a volunteer
who, by his efforts exerts a kind of personal leadership in that area.
I think that is a pretty good summary.
yours
Martin.
