On Wed, 27 Feb 2013 11:28:34 +1300, Matt Madia <mattmadia@xxxxxxxxx> wrote:
Yes and lets start R2 now, waiting for R1 to finish just stalls the wholeproject. I think the R1 BeOS compatibility goal served the project very well to get to a certain point but now it just slows it down I'm for an official R2 developer branch. That would give developers who are more interested in new features a better platform.IMO, we should have an idea of what Haiku will and will not become in R2, *prior* to creating its development branch. Is it simply for breaking compatibility, so as to implement the API and features that we want --or-- is it going to be something of a much larger scale? What measuring stick will we use to decide if some commit is or is not appropriate?
yes good point and I kind of anticipated that comment :-)Yes there should be some general consensus what features are desirable. But I think that is not a big problem. The goal of this branch should actually be that people can try stuff. A R2 dev branch nightly would expose this ideas to others and help to find design flaws. I think a certain bit of tolerance from both sides would be needed: on one hand people are allowed to try something new, on the other hand they have to accept the fact that their work may be removed again. Furthermore, compatibility to R1, if possible, should always kept in mind.
An example for such a feature with general consensus (I think) is the session manager. But because it was > R1 I lost interested, as a result this also left the restoring and saving of S&T groups implementation in a very bad shape.
To summaries, would propose a common dev branch that is kind of moderated, (via a this mailing list). Or to say it in a nerdy way: we have to remove the global R1 development lock to release all the waiting developers! ;-)
Regards,
Clemens
