On Dec 20, 2007, at 9:15 AM, Garrett D'Amore wrote: > You're missing a very simple and obvious point though. Right now > whatever the community votes on, if you don't have CTeam approval > (a Sun entity), you won't get integrated. This is because Sun owns > the code.
Ownership of the code is irrelevant aside from licensing. Sun says we own the process. I am well aware that folks inside Sun still think that CTeams rule the roost. That is the point. The OGB, unlike the CAB, is not an advisory body. CTeams no longer have the power to which you speak, and the only reason they still exist is because no other power has supplanted them. The OGB needs to use its power to enforce the community rules. If the community decides it doesn't like the result, then the community can change those rules. Right now we have rules that nobody follows, and that is the worst possible situation. > I believe that the folks that put in place the rule requiring CTeam > approval are at least as senior as those who signed the charter for > OpenSolaris. Unless they are the CEO, no, and in any case they would have to explicitly rescind the charter. That is the whole reason we created the charter -- so that we would know what the boundaries are and how much Sun is willing to let the community decide. > So, there are two conflicting policies here. Actually, only one policy exists and it is conflicting with an old tradition. Something has to change for Sun to be able to take advantage of the open development process. Sun made that choice. And yet we still expect it not to be true. Why? > Until OpenSolaris has its own mercurial repo, it will be hard for > votes by community groups to integrate code into Nevada to carry > any weight when those votes are not aligned with Sun's interests. > > One could argue that we (OpenSolaris ON community) could create a > repository of its own, apart from Sun's. But that hasn't happened > yet (and I think there are good reasons for that.) The device > driver community itself has no consolidation nor repository of its > own, so a vote within the driver community could only grant commit > permission to a repository that doesn't exist. That sounds like an excellent summary of a high priority problem that the OGB can solve before the end of this term. Right? ....Roy
