The starting point for this discussion centred on the development of core libraries that the Qpid/ActiveMQ projects could use for 1-0. (Clearly I'm interested in how this might work for the Qpid implementation of AMQP.)
This is a great idea and would be really useful for both projects on the Apache side. How many Apache projects are dependent on Google Code hosted libs for their core implementations ? I honestly have no idea but I think thats an important question for the Apache teams who will want to implement 1-0 in their brokers. Is that not what you meant to suggest would be the end result Bruce ? If the objective is mainly about having a free/loose collaboration of individuals/organisations on an AMQP core libs then thats a different question, I think, and the detail is less important. Marnie On Fri, Jun 11, 2010 at 2:01 PM, Rajith Attapattu <rajit...@gmail.com>wrote: > Frankly I am a bit baffled about the concerns raised here. > > Rob, Bruce, Rafi and a few individuals have expressed a desire to > start an amqp protocol client project that is independent of any AMQP > projects that already exist. > The goal is to enable collaboration from anybody who is interested in > AMQP and to avoid duplication of effort. > To this end Bruce has extended an invitation to the developers in the > Qpid/ActiveMQ communities to participate in this. > > **Also as Rob mentioned, it's upto the Qpid/ActiveMQ communities to > use these client libs or not.** - However common sense suggests that > we stand to be benefit if we use it, provided that it's designed > properly which I am sure it will be. > > If we think pragmatically, there are many good reasons why such a > project needs to be independent instead of being affiliated with an > existing AMQP project. > > 1. All though Qpid, ActiveMQ, RabbitMQ etc.. are open source projects, > they all have their own identity, their own communities, different > companies backing them etc. > If this proposed project goes under any of the existing projects > then it is inevitable that people will feel that the chosen community > will have more influence over the project than the others. > This is not a good thing ! > > 2. If this project is independent, then we will likely get > participation from a wider audience, not just folks from the above > mentioned communities. > > 3. It is IMO absurd to require a RabbitMQ developer to earn karma in > the Qpid/ActiveMQ project to contribute effectively to this project. > > Regards, > > Rajith > > On Fri, Jun 11, 2010 at 4:04 AM, Robert Godfrey <rob.j.godf...@gmail.com> > wrote: > > On 11 June 2010 08:48, Bruce Snyder <bruce.sny...@gmail.com> wrote: > >> On Thu, Jun 10, 2010 at 8:54 PM, Marnie McCormack > >> <marnie.mccorm...@googlemail.com> wrote: > >>> I'm not sure what the best solution is, but I do know that I'm missing > what > >>> the problem using one of the Apache projects for this work is ? > >>> > >>> It seems like there some context missing from this debate, possibly > from > >>> discussions at the F2F. > >>> > >>> Bruce - can you elaborate on why other people might not want to work on > AMQP > >>> under an Apache project but would want to work on it as another open > source > >>> project and use an ASL license ? > >> > >> Why would someone who works on an AMQP broker implementation that is > >> not Qpid want to come to the Qpid project to contribute to an AMQP > >> protocol level client? If this project were hosted as a subproject to > >> Qpid, it would forever be associated with Qpid instead of being > >> neutral from any broker. > > > > Moreover those people would not find it possible to commit without > > gaining committership to the relevant Apache project. I was talking > > this evening to one of the guys from RabbitMQ who would very much want > > to be involved in this - and I very much want to encourage this. > > > > I think we should see this idea as something that may be of interest > > to particular Qpid, ActiveMQ, Rabbit or other developers... not > > something that involves those projects in themselves. In the future > > those projects may decide to adopt work that is done in this space (or > > not)... but really what we are saying is that we feel like there are a > > number of people from different communities who feel like it would be > > worthwhile to have a go at this - and we're seeing how many other like > > minded folks there are. There are, of course, also likely to be > > developers in these communities to whom this is of no interest > > whatsoever. > > > >> > >>> It seems a little like double overhead for those already working on > Qpid or > >>> ActiveMQ to have a third project in the loop, with different lists and > >>> process and all that. Having been around for a while I understand the > >>> overhead with Apache, but I'm also concerned that we haven't really > talked > >>> about what we gain from a non-Apache project - what the gain from doing > the > >>> core libraries somewhere else ? > >> > >> Because it lowers the barrier of participation and is not associated > >> with any particular AMQP impl. > >> > > > > +1 I think it was a great sign that so many people from different > > communities were willing to try to start up something like this - and > > obviously if there are more people out there that will be even > > better... I really want to do as much as possible to encourage this > > effort and lower the barriers of entry for anyone who is interested as > > much as possible > > > > -- Rob > > > > --------------------------------------------------------------------- > > Apache Qpid - AMQP Messaging Implementation > > Project: http://qpid.apache.org > > Use/Interact: mailto:dev-subscr...@qpid.apache.org > > > > > > > > -- > Regards, > > Rajith Attapattu > Red Hat > http://rajith.2rlabs.com/ > > --------------------------------------------------------------------- > Apache Qpid - AMQP Messaging Implementation > Project: http://qpid.apache.org > Use/Interact: mailto:dev-subscr...@qpid.apache.org > >
