[OpenLayers-Dev] time-based versus functionality-based release

[Christopher Schmidt]

On Wed, Nov 05, 2008 at 09:17:56AM +0100, bartvde at osgis.nl wrote:
> Hi Chris,
> 
> you've got a good point there.
> 
> My "theory" is that momentum gets less after a release, so maybe if the
> release period had been longer, those features would have made it into
> trunk/2.7 already. But again this is purely speculation.

I see a lot of motivation that we never do many things until someone
pulls the switch and says "Okay, we're going to do a release!" In the
past, I was taking a more active role in reviewing/ applying patches --
I spent every weekend doing that. Since 2.6, I have backed off of that,
and I think the current state of the tickets marked for 2.8 shows that.

Without an external party going through and reviewing ticekts, things
don't move forward. I'm not convinced that release deadline helps that,
because the only way that release deadlines seem to be successful is if
someone (who isn't writing new code) says "I'm going to do the work to
make sure $x gets in" -- and if they're wiling to say that, it seems
like they should be willing to say that without a release just as
much...

Anyway, I think that this is evidence that the work of organizing
reviews of existing code is a bit stalled at the moment, and I'm not
convinced that delaying the release would have changed that. (And if
the WFS protocol gets into trunk, I'm 100% behind the idea of cutting a
new release against trunk, and i'll even do the legwork: it's time to
moev on from Layer.WFS.)

Regards,
-- 
Christopher Schmidt
MetaCarta