On 7/5/2012 9:31 AM, Benson Margulies wrote: > A thread on the subject of EasyAnt highlighted some perplexity for me, > and I think that it deserves a thread of its own. > > Once upon a time, the Foundation had a set of 'Umbrella projects'. > These were PMCs that managed a collection of 'sub-projects', each > considered a relatively independent community. > > A few years ago, the board concluded that this wasn't working, and > asked the umbrella projects to fission. At this time, that process is > pretty nearly complete. > > It seems to me that the concept of 'project sponsorship' of podlings > is left over from the days of umbrellas -- a podling sponsored by an > umbrella would graduate as a subproject of the umbrella. > > If this logic is correct, then it would seem that project sponsorship > is no longer relevant, and we could remove it from documentation. > > Either a podling is going to become a TLP, in which case project > sponsorship is irrelevant, or it's going to merge into some existing > TLP, in which case I claim that the TLP could handle the process > without the overhead of a podling at all. We have an IP clearance > process, and TLP's are free to create sandboxes in source control and > grant commit karma to them as a means of integrating an incoming > group. One use-case for a podling that might be destined for an existing TLP: if the podling needs to grow community/committers, and/or learn how to work in the Apache way, then living in the incubator may tend to these needs better than joining a TLP. This may be particularly the case where the podling is implementing some new thing that the TLP project's existing committers may not have enough time/interest in to dig into the code, although they could see it as a valuable addition.
-Marshall Schor --------------------------------------------------------------------- To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
