[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Now for something really controversial...

> 	Christian Perrier (bubulle) always says that, sometimes,
> translators face a hard time exchaging between several different VCS
> because each maintainer has its own preferences, so at least the basic
> commands are required to help them. :-)

I'm even more direct than this.

Distributed VCS are of no added value for localization work. 

They make the work way much more complicated. This is mostly because
the files translators work on actually have their content *depending*
on the files modified by developers. That's actually the case for
software translation where the translation material depends on the
course code.

So, the need for a "central" reference point is vital. Which makes the
major enhancement of distributed VCS pointless..:-)

Another argument against distributed VCS is the higher learning curve.

While I can describe basic SVN operations in a few words (see
http://people.debian.org/~bubulle/d-i/i18n-doc appendix), I never felt
this possible for distributed VCS.

Of course, I perfectly understand the interest of distributed VCS for
developers (maintaining branches, working without network access,
etc). But my point is that all repositories I have to interact with
which are distributed ones (bzr for APT, darcs for aptitude....tla,
then baz for dpkg in the past) have been a major PITA for me.

For instance, I spent like 3 hours recently unscrewing my local copy
of APT which is used by the main maintainer as a reference for all
translation material. And this was with the help of bzr wizards like
Otavio Salvador or Colin Watson....all alone, I should gave given up.

In three words: huge learning step.

Attachment: signature.asc
Description: Digital signature

Reply to: