[Two weeks for gamma rather than one]
Sure.

[changing the plan for 3.7]
We could. We might want to consider it after we're done with 3.6.

[opening alpha before beta ends]
I'm against - I think the focus on just bug fixes is a good thing, we
should make the most of the beta period. I don't think we're still at
the point where we can call 3 weeks serious rot - I suspect the average
life of stuff in the BFA is still much much longer. Thought it is
getting better.

Daniel

Doug Way <dway at riskmetrics.com> wrote:
> 
> On Thursday, July 24, 2003, at 04:09 PM, Daniel Vainsencher wrote:
> 
> > I propose we declare first Thursday of September (5.9.2003) our new
> > release target, with august 29th for the move to gamma, giving us an
> > additional month of beta. Considering that we're still doing reworks on
> > contexts and the network rewrite, I think that's reasonable.
> >
> > A lot of base stuff has improved in this version, it won't hurt us to
> > take a little more time to stabilize it. It will also give more time 
> > for
> > people to actually switch to be using the beta, so we'll have more
> > people testing it.
> 
> This sounds like a good idea.  Although the original plan called for 2 
> weeks for gamma, which I think would still probably be a good balance, 
> since we're stretching out the beta by quite a bit.  So, that would be 
> August 22nd for gamma.  (although I don't feel really strongly about 
> this)
> 
> Hmm, if we postpone the gamma this far, we may want to open 3.7alpha 
> earlier than the 22nd/29th so that the various enhancements that have 
> been posted don't "rot" for too long.  Although I suppose they won't 
> really rot that much since the fixes going into the beta image are 
> typically much smaller changes than the stuff that goes into an alpha 
> image.
> 
> Also, we might consider whether we want to adjust the schedule for 
> 3.7beta/gamma/final to be more like what we're doing for 3.6.
> 
> - Doug
> 
> 
> > Daniel
> >
> > Tim Rowledge <tim at sumeru.stanford.edu> wrote:
> >> I really think we need more time for testing and fixing. At least a
> >> couple of weeks extra would be smart.
> >>
> >> tim
> >> --
> >> Tim Rowledge, tim at sumeru.stanford.edu, http://sumeru.stanford.edu/tim
> >> Useful random insult:- Several nuts over fruitcake minimum.