> It's not a question of what you *think* you can maintain. If you bring an > external code base into the ASF repository, you *must be committed* - as a > community, not one or two individuals - to maintaining it and resolving > any technical issues, not to mention resolving any legal issues *before* > it touches the ASF repository. If there is even a hint that this is not > the case, you should go through the incubator.
In the case of Jesse's SF code it doesn't sound like incubator is required. This is based on Craig's detailed explanation as well as common sense. According to Craig it is fine for a few individuals to work off list on things and then offer them to the community for inclusion. Its up to the PMC to decide whether or not that code should then go in the sandbox or could be added directly to the source code proper. Also according to Craig, it sounds like its equally fine to develop the code inside a sandbox and have those doing the contributions rely on existing committers to contribute patches, etc. This is my strong preference. Its a bit of a hassle but it will be easier for the entire community to participate if the components are in one central location. When a component looks to be ready we just vote it out of the sandbox. There can be a clear line of delineation between what is officially sanctioned MyFaces code and what is still "experimental." > Please don't think of this as "can we avoid doing things the right way and > just stick it in our repo?". Intended or not, that is how this thread is > coming across, at least to me. The incubator is there for good reasons, > and, as Ted has pointed out several times, it need not be a heavyweight > process. If something is brought into the incubator, and it's in good > shape with a community forming around it quickly, then there is no reason > for it to remain there for a protracted period before moving out and > becoming part of an existing project. > > One other thought I'd like to throw out here: As it is today, the MyFaces > community is faurly small, and relatively inexperienced in the "ways of > Apache", albeit with a few mentors keeping tabs on it. Given that, I would > suggest focussing on what's here now, and grokking the ways of the ASF, > before rushing into how to accummulate more code from other places. I agree that we are relatively new to the ways of Apache (at least from a rules and regulations standpoint.) I don't think we are really focused on accumulating code from other places. I think those of us proposing a sandbox are doing so as a way to study code that come from *within* the community but that still needs to be vetted before formal inclusion. To me that seems entirely appropriate (and a good way of encouraging potential new committers to contribute.) > Martin Cooper sean
