Miloslav Trmac wrote:
Hello,
On Sat, Sep 27, 2003 at 10:30:45PM +0200, Nicolas Mailhot wrote:
>We are not talking about writing software from scratch here. To get into
>Fedora a project must at least reached the packageable state. If at this
>stage the project maintainers haven't even started thinking about
>localization we should really worry because there's little chance it
>will be done by the time Fedora core ships.
Speaking as a translator, I don't think adding translations to
_existing_ software projects just for Fedora is right. The translations
should be done, integrated and tested upstream, so that they receive
wider testing, developer cooperation and so that they benefit all
users of the package, not only those using Fedora.
Sure.
However localization should be a criterium for acceptance in Fedora, and
Fedora will be a good testing ground for upstream translations.
It's much easier to ask testers to review translations in packaged
software than asking them to grok the english readme and install stuff
by themselves. I don't believe in closed room localization. Real users
will find real problems in localization just like they will find some in
software code.
That the reports then need to be filtered and fixed upstream is not
specific to localization at all. I strongly believe Fedora will largely
live or die by the way such reports are handled. A lot of people are
dying to report problems but do not have time to invest to learn all the
various reporting systems upstream projects use. If they could report
everything via a common interface then have people in the know collect
reports and interface with upstream projects overall distribution
quality would improve enormously (and I'm not just talking about
localization here).
Cheers,
--
Nicolas Mailhot