Open Source Message Queuing
"The formation of the Working Group follows a multi-year development effort at JPMorgan Chase aimed at creating an alternative to expensive proprietary messaging middleware such as IBM's MQSeries. "
See also:James Governor's MonkChips
Q: There are many folks on Wall Street that are working on an open source equivalent of MQ Series. Would be curious to know what industry analysts think of this effort? Would also love to know if it possible for a traditional end-user client of an analyst firm to create software and get it on a quadrant without having to become a software vendor?My emphasis added
A: I also talked to Coridan, which is another open source message queing effort, as per your earlier suggestion, and while I thought the MantaRay approach was potentially interesting, I am a little worried about who the target market is in both cases. Big Wall Street firms have casts of thousands that love to build IT stuff. They have people that like to go toe to toe with IBM and other major vendors, architecture for architecture, flexing technical pectorals. But the level of resource in financial services is hardly the enterprise norm. Banking is IBMland: it makes sense to put pressure on it. But MQSeries, now called WebSphere MQ, is pretty darned solid middleware. I do wonder sometimes about the balance between function and "non-proprietary" choices. Many SOA implementations are driving WebSphereMQ volumes, without requiring other WebSphere componentry. It is a very handy transport which covers an unbelievable number of platforms. It scales in a way JMS can't. Without a vendor sugardaddy on board, what happens if banks cut the AMQ budget? What is the maintenance and governance model? I am all for enterprises contributing to open source, and it will be interesting to see if AMQ becomes, say, an Apache submission. Ron at Zapthink talks about open source message queing putting pressure on ESBs. I think of MQ as a transport not an ESB. In the end, I think I need to talk to AMQ and JP MorganChase to find out more. I have many questions before I can really say what I think of the effort.
AMQ might even act as leverage to push IBM towards thoughts of open sourcing the core WebSphereMQ messaging service. Its not as if customers wouldn't still come to IBM for service and support and the rest of the stack... and who knows, IBM might even find one of its venerable middleware platforms winning new customers by lowering barriers to participation. IBM could then focus on selling the higher level software, for SOA Governance and so on. I wonder what the guys at Hursley think, not that they will likely say in public...