On 28/04/07, Thorsten Scherler <[EMAIL PROTECTED]> wrote:
to have a quick release let us start early to plan and have a well defined feature list.
I agree in principle, but I'm not sure how active I will be on 0.9. We are using Forrest again (hurrah!), but there are issues with it that may see us going in a different direction as the project matures (see my original RT thread prior to 0.8 relase). I'll be addressing these issues by resurecting my RT once I have come good on my promise for a release management proposal. This proposal will incorporate a proposal for parallel development on 1.0 and 2.0.
I think this three points can be done before the end of the year where we should release 0.9.
Bahh!!! 6 months is too long. We should be aiming for 0.9 in 3 months tops. The Ivy build will not take long, we can then do integration testing on updated jars, bug fixes from 0.8 and issues we don't want to roll over again (see Davids mail in this thread) will be sufficient for a 0.9 release. We need a much faster release cycle every three months at least. Don't worry I'm going to be proposing a more streamlined release process which includes: - more automation - more automated testing - stable and unstable releases (just like Linux Kernel or Apache HTTPD) One more note, dispatcher is a plugin so it has an independent lifecycle and 0.9 should not be dependent on it. I am in favour of it coming out of whiteboard but not the 0.9 release being dependent on that happening. Ross
