[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-i18n-doc
Subject:    Re: Red Card for Polish translation team
From:       Nicolas Goutte <nicolasg () snafu ! de>
Date:       2006-07-15 8:33:16
Message-ID: 200607151033.16706.nicolasg () snafu ! de
[Download RAW message or body]

On Monday 03 July 2006 04:02, Abel Cheung wrote:
> On 7/3/06, Mark Kretschmann <markey@web.de> wrote:
> > > I find the cards system quite harsh (and a bit silly), but that's only
> > > my opinion. And maybe I'm wrong because the Polish team was aware of
> > > the problem but simply ignored it.
> >
> > It is harsh, but we think this may be a chance for the affected
> > translation team to rethink their quality assurance. Ultimately the user
> > will benefit from it. As mentioned, this sort of error has hit us
> > regularly on recent releases.
>
> This could have been avoided in the first place. I'm speaking of
> CVSROOT/commitinfo file in CVS system, but not sure if equivalent
> feature exists in subversion. CVS can be set up to behave like,
> whenever a po file is checked in, a script is automatically run
> to check if the po file is valid. I think that can eliminate
> the need of such whatever card system.

This is possible in SVN too, by using the pre-commit hook. However checking 
the validity of PO files is probably too much runtime consuming, 
unfortunately. (KDE SVN has much more "traffic" that the Subversion SVN 
repository, which uses this.)

That is probably why it is not implemented in KDE SVN.

>
> Abel

Have  a nice day!

>
> > At some point we might set up an automated build bot for Amarok that
> > catches issues like these immediately. However, until such a system
> > exists, we developers suffer from any mistakes like these, and also the
> > users. For example, in this case the (silly) mistake meant a whole day of
> > delay for distribution packages.
> >
> > --
> > Mark
[prev in list] [next in list] [prev in thread] [next in thread]