While I agree that it's important to confront the current inventory, I think that the situation at hand is as much a matter of what goes into the incubator as what comes out.
In my limited experience, size and growth are a big barrier to graduation. If a project is chugging along exhibiting good Apache process, but not quite achieving the necessary size, it poses a dilemma. No one wants to exile them to the outer darkness, and no one wants to graduate them, either. Perhaps one implication of this is that the bar for entry with respect to community size and diversity should be higher. In the modern world of (ahem) github, no one needs Apache infrastructure to run an open source project. A small group of people have any number of choices for free and open hosting. They can come to the incubator when and if they get are big enough to have a strong likelihood of success. I can see two problems with this view to begin with. One is IP management. The more people participate in a project and the longer they do so, the messier it becomes to track provenance and get ICLAs from all of the contributors. One almost wishes that the ASF could accept ICLAs and code grants for code that is in fact living somewhere else. I'm sure someone will tell me why that idea is nuts. The second is that growth is not cured by a minimum size. We want to see a podling do a good job of accepting new contributors. Even if a group of 27 people shows up on a proposal and commit their brains out, a year later there may still be 27 of them. I think that a solution here might be for such projects to be graduated but to have a requirement of a PMC member, or even PMC chair, who is an iPMC member and who thus can provide an extra level of detailed supervision. --------------------------------------------------------------------- To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
