I'll try to refocus this discussion on a more requirements basis. I believe there are two real questions to answer, about any Language X project for libcloud.
1) Are there enough people to form a cohesive community for development and maintenance for a libcloud port to Language X? 2) Ideally, can language X reuse implementations or entire parts of the existing code from libcloud to make it easier to develop and maintain? I am happy to consider other questions, but fundamentally, for any Apache project, the first question is the only one that matters -- are there enough people interested in a project to review the code, vote on the releases, and build a strong community around the code. If there is only 1 person, then the project should just go to Google Code or Github, and thats fine, thats just how many open source projects are -- Apache is about building a community around the code, it doesn't really matter what the code is, where it comes from, who wrote it, or even if there is another 'competitor' within Apache, or in the wider open source world. On the specific of topic of Libcloud for Java, I think the answer to the first question is yes; There are many java developers in the Apache ecosystem in general, and several have expressed interest in this thread to get started -- before any community has had time to be built. I think thats a good enough sign to 'try it'. I think we should consider deeply if there is a way to re-use the existing python code to power a java version. I think this brings many benifits to both sides, like wider exposure to libcloud in other venues, and hopefully makes building and maintaining the java version much easier. I think there are many possible ways to do this, including embeding Jython, using an RPC mechnisim like Avro/Thrift or even just JSON and HTTP. I think those are the interesting technical discussions to have -- a final decisions on where the code lives, or what PMC manages it is largely irrelevant, it will get sorted out as the community is built, and if they turn out to be fundamentally different sets of people, there is much precedent within apache for projects to give birth to new projects. Thoughts? Thanks, Paul
