On Apr 7, 2011, at 9:17 AM, Stéphane Ducasse wrote:
I do not have specific stance on that this is why I hear both arguments.
Now agility is not something that we claim but that we do. I think that not having releases that drag on forever is important. I prefer to have - 3/4 releases full of energy that are time based vs. 2 that are slugglish and ends up in few fixes over a long period. - this would avoid to have an accumulation of code to integrate and laucnh beta too early to reduce the pressure. - In addition since people do not understand what is a release candidate, then having more frequent release would mean that people port more often if they want. So we will see. For now the good point is that we can release 1.3 tomorrow and this will not be a release maintenance because lot of changes are already done.
The next steps are: -> go over the bug tracker and tag as 1.3 just the real important entries -> Fix the easy to fix failing tests (just 25 to go) -> Do another iteration on the build server for complete auto-build to the artififact we ship. and: -> Do another iteration on Cog builds so we can move to Cog only. I want a 1.3 that is in "we could release today if we wanted" 90% of the time. -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.