I'm encouraged that the overhead for opening an incubator is less than I had thought, so I'll begin drafting a proposal for an incubator. Thanks to all for the help and advice and putting me straight, and I look forward to your continued help over on the incubator list once I have a first draft proposal in place.
Kelvin. On 05/01/2008, William A. Rowe, Jr. <[EMAIL PROTECTED]> wrote: > > >> On Jan 4, 2008 1:45 PM, kelvin goodson <[EMAIL PROTECTED]> > wrote: > >>> We'd like to propose the creation of a new lab for the purpose of > >>> developing the reference implementation and technology compliance kit > >>> for Java... The reference > >>> implementation would be seeded by a contribution of code from BEA, > > Ok - that's straight out AFAICT; seeded external code must pass through > the incubator anyways, in any case. Labs seems entirely inappropriate > as a starting point (although like any other TLP, labs can vote to > refer and recommend this to incubator). > > If you want to fast track that, you can pass this through incubator > for IP Clearance only to pull it into the foundation, accepting it back > into Labs. But incubator is already in the loop. > > >>> the TCK would be seeded from the community test suite developed > >>> in Apache Tuscany > > Putting together a first-draft of the TK seems to be fine; you are forking > already-OK code from within the foundation. > > kelvin goodson wrote: > > The thinking is that the lab would allow us to be agile at this early > stage > > as we'd like to deal with the issues of community overlap separately at > a > > slightly later stage. > > Sometimes that worries me; it implies that incubator doesn't provide any > agility from the start; that's wrong. You can't release the RI or the TK > from the labs anyways; why not develop in a podling and start with only > three programmers? If that grows during the initial phase, great, but I'm > not sure why you believe that the incubator is going to hamper agility? > > The labs will hamper on the side of bringing to completion, and then it's > back to incubator. Incubator won't graduate it without a community, and > by starting there, you can demonstrate growth over time. > > I agree that writing proof of a concept, labs can be helpful. But the > concepts already exist in BEA's code, in Tuscany, and in the JSR, so > making > a case to the incubator doesn't require a whole lot of preparation. > > Bill > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > >
