On Sun, 2005-07-24 at 00:30 -0700, Bryce Harrington wrote:
Hi all,
As the release of 0.42 nears, I haven't been able to help thinking about what we should do for the next release. I'd like to throw some thoughts out for comment.
I think we're all ready for a bit faster release cycle. 0.43 really took a bit too long. Partly that's because we wanted a nice long development cycle, but mostly because we had some extremely ambitious goals.
Agree...
This release, though, the goals are more modest (and already mostly done). As well, we have four Google Summer of Code students producing code, and I think it would serve their learning process (and our own ability to review the code and give them feedback) if they can start getting their code drops out to the Inkscape audience.
Yes, I think this is imperative for them to get full exposure to the process. Plus, since we tentatively know that Google is planning on doing a Summer of Code in the Winter (So, winter of code?), then this will set up a good precedent for this type of power dev.
Thus, I'd like to propose that for at least 0.43 and 0.44 that we shoot for something on the order of a month each:
3 weeks development 2 weeks feature freeze - stabilization / bugfix ~1 week hard freeze release repeat
As far as specific development goals, I think we should just set the one goal of having some working code from each of the SoC students, (plus whatever else gets done by others). Some of the students may need our assistance in learning how to integrate code, test it, fix bugs, and so forth.
(The originally planned goal of the 0.43 release was to work out how to do extensions, which thanks to hard work by Ted, ACSpike, and others, is already fairly well in the bag.)
Anyway, let me know your thoughts on the above, and if it sounds acceptable, we can start on this plan of action as soon as the release of 0.42 is complete! :-)
I think this sounds quite good. I think in the end we will work towards whatever plan we design. Thus, if we allocate 4 months to a release, we will probably do it in 5. And similarly, if we allocate 1 month, we will do it 1.5. Thus, I agree that we should try to get 1-1.5 month releases for a while. With this release we have verged on releasing to late and not often enough.
The plan you set forth sounds good. With regards to extensions, it would be nice to get an update about this, what is still needed, etc. I'm a little lost to what the state is of this dev.
Bryce, are you going to update the roadmap?
Jon