Artenaut End of world schrieb: > I know what you are trying to tell me. Perhaps, > I explain wrong. I try with an example: > > - 2006: Cinelerra-HV 2.1 > - 2006: Cinelerra-CV 2.1 (merge Cin-HV 2.1) > - 2009: Cinelerra-HV 4.1 > - 2009: Cinelerra-SV 2.1 (why not merge into Cin-CV 2.2??) > - 2009: Cinecuttie 1.0 (why not merge into Cin-CV 2.3??) > > Who decided what to merge into CV? No one. There is no one really in charge. In former days, the CV-Community was defined by "doers", but as long as I can recall, there was no more formalised responsibility or some procedure how to come to decisions.
You need to understand, the greatest difficulty with Cin-CV is the relation to "upstream". In the best times, when Cin-CV had a rather active group of core hackers, the regiment regarding additions and reworks was rather rigid, but OTOH this resulted in a sequence of well defined, isolated patches. Some of them where sent to "upstream" and a subset of these actually re-appeared in the next "upstream" version (from HV). But basically the problem is: whenever you touch something, you're creating yourself problems with the next merge from "upstream". And who is going to resolve these? To add to this: whenever you try to fix or change something in a more disruptive manner, i.e. trying to resolve the root of some problem, the chances that anyone (aside of yourself) wants to merge the next upstream version with your work, rapidly go to zero. Another thing worth mentioning is that for most users of Cin-CV, the application basically works satisfactory. You've learned to circumvent the problems and can do your work. No new excitement, but also no new bugs. Everything what came from "upstream" in the last versions always caused a lot of additional problems, which had then to be fixed by the active community. (For sake of completenes I should add that also the CV-Version added some problematic things which aren't resolved yet). Currently there aren't too much active Cin-CV hackers left which really know the code, thus any merge or change at the half-way stable version 2.1 is risky. Thus the question really is: What to do, how to proceed into the future?? I think, a round of brainstorming and discussion about that topic would be desirable, indeed. Just rationally speaking, *if* we want at all to merge in the HV versions, then the next thing to do would be to merge 4.1. Merging any new experimental features in a 2.2 or 2.3 as you proposed would just be a waste of efforts. Because you can expect the following: after merging 4.1 -> Cin-CV, chances are that a considerable part of these experimental extensions would have to be re-done or at least re-worked and adjusted, before you could even think of accepting or rejecting these extensions into Cin-CV. Probably you're now able to understand better, why we're not "behaving ourselves".... Cheers, Hermann V. _______________________________________________ Cinelerra mailing list Cinelerra@skolelinux.no https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra
