Re: [Gnash-dev] Gnash 0.8.9 status

[olafBuddenhagen]

Hi,

On Fri, Mar 04, 2011 at 09:18:14AM +0100, Sandro Santilli wrote:

> If there's a release goal, that goal blocks a release too.

To be honest, I don't think explicit release goals are a good idea in
general. New releases should happen as soon an there are any major
feature improvements. (And no major new problems introduced.)

> The main reason why I'm happy to have a 0.8.9 coming out is that 0.8.8
> had a couple of regressions over 0.8.7, making it unpleasant to play
> winterbells [1] :>

Indeed, regressions are always problematic -- a new release should never
(knowingly) make things worse for users... In fact, I'd consider
"regression" to be the major criterion for release blockers. If new
functionality is buggy, it's usually still better than not having the
functionality, so a release is still an improvement for users...

> Seriously, this new release I see more as an improved-stability one
> than anything else,

Well, 0.8.9 also brings considerable performance and correctness
improvements for video playback -- which probably affects most users
even more than the stability improvements :-)

> Nothing prevents us to ship a 0.8.10 in a few months

Indeed, I hope to see the next release pretty soon :-) As OpenVG is
scheduled to land shortly after 0.8.9, there is already a candidate for
triggering this...

> So, is your candidate blocker in the Important set already ?

I didn't have any specific bugs in mind. Just wanted to bring it up, as
Benjamin's mail made it sound as if "goes up in smoke" was the only
criterion for a release blocker :-)

-antrik-