Re: Typos in eog.schemas.in

[Mark McLoughlin <mark skynet ie>]

Hey,

On Tue, 2004-02-17 at 22:48, Danilo Segan wrote:
> Hi Kjartan,
> 
> Kjartan Maraas <kmaraas@broadpark.no> writes:
> 
> > My opinion on breaking string freeze is as follows:
> >
> > - no changes to non-visible strings (schemas, atk descriptions etc)
> > - approval needed for changing strings visible in the GUI (still not
> > recommended to do so of course)
> > - marking existing strings for translation is ok (this is probably the
> > most controversial one to some)
> >
> > How do others feel about this?
> 
> I pretty much agree with you, though not unconditionally on the last,
> "controversial" point :).  I'd be for something like: it's ok to mark
> accidentaly left unmarked strings in the first week (or whatever is
> the period you called 'soft freeze' earlier), and after that, they
> need approval as well.
> 
> As for the GUI strings, I'd still hardly oppose any typo fixes —
> we've got two choices: either leave a mistake in one English message,
> or submit >fourty translators to fixing all of their translations,
> with the risk of some not being updated in time.  So, perhaps there's
> place for soft freeze here as well — in that period, typo fixes are
> allowed, after that, only in very specific circumstances.

	One thing I would recommend is that you guys come to a decision on a
very clear and simple policy which you and the release-team can
implement. A vague, flexible policy with many grey areas is probably
just going to cause even more confusion amongst developers and you'll
likely see even more freeze breaks.

	Also, I keep seeing a pattern that when it comes to this time of the
release, there generally ends up being a discussion about what a "string
freeze" means exactly. It would be good to get the conclusion from this
documented so we don't have the same discussion coming up to future
releases.

> Of course, it'd probably be nice to have some kind of approval like
> release team does ("two approvals is good enough", with probably a
> higher number of required votes), without having to depend on
> Christian Rose being available at all times (not that I have any
> problems with his decisions; he may sometime be unavailable just like
> a week or so ago, and we as a community want to keep going during
> that time).

	Yeah, redundancy is always good. Pick a number of people on gnome-i18n
who understand the policy and its implications and bless them as
approvers. A single approval should probably be enough, though.

Cheers,
Mark.