Re: How frozen? (Re: Patch: use std::string)

> (a) we need some feedback on the AbiCollab stuff, which is the principal
> new feature for the 2.6, and the Win/Mac UIs for it still need to be
> created based on that feedback.

I feel that this is simply unrealistic, given our present condition.
The MacOS bugs targeted @ 2.4 and 2.6 suggest that it'll take another
year for it to approach parity with our Win32 port, let alone the Unix
one. Given our relative lack of resources in the MacOS and Win32
communities, I'm ok with stragglers, absent a realistic promise to set
our affairs in order. "Let's wait it out and hope for the best" is an
unacceptable plan.

> (b) we need some feedback on the Pango rendering from a non-Latin world.
> (I am uneasy about the OLPC using the 2.5.0 in this regard, but that is
> their choice.)

Agreed, but then we need to be more proactive about things if we're
going to consider this to be a show-stopper. We can't say "we'll let
it lie around for a few months and hope someone tries it with Indic
and then reports it in bugzilla." Especially since our 2.5.0 Unix and
Source download stats are presently a rounding error, mostly come from
western nations, and people's general disdain for bug trackers.

> I think we need to make sure that both of these new features work well,
> and are rock stable to avoid returning back to the 'AbiWord misbehaves
> all the time' reputation that got in the 2.2 series and worked hard (and
> successfully) to get rid of in the 2.4 cycle.

That's why I set my bar at "as good as or better than 2.4.latest" for
releasing 2.6, and not "leaps and bounds better than 2.4.latest".
Perhaps my standards are too low, and that is why I said "ok, maybe a
month isn't enough time." It's ok to change one's opinion when
presented with new evidence and new arguments.

Cheers,
Dom

-- 
Counting bodies like sheep to the rhythm of the war drums.