Friday, January 4, 2008

CEP vendors: RIP Kaskad, new Pion, and same IBM

End of 2007 has been an interesting time for the Complex Event Processing (CEP) ecosystem.
First, Kaskad adds itself to the RIP list and closes business as reported by Marc and confirmed by Kaskad former CTO Colin Clark. This was 16 employees. So for those who would argue that a company with 16 employees is going to sustain better than one with less than 5 in the same market, well, not that obvious.
I am not surprised by the news. Back in February 2007, Kaskad had announced it would open source part of it offering. I had asked them for more details both privately and publicly several time as I found the news interesting especially from an Esper and EsperTech perspective but never got any answer from Kaskad team. I wonder if they weren't already dead back that time...

Other news I could spot thanks to Marco: some Californians have started to chase the CEP market with an open source approach ala Esper - Pion from AtomicLabs: Our goal is to build the world's #1 open source platform for complex event processing.
It is actually quite far to compare to Esper, might be implemented in C++ and not Java / .Net, and has yet to be proven and most importantly has yet to expose some real CEP features especially some Event Processing Language (EPL). Data or event processing is not CEP. Welcome aboard guys and good license choice by the way (Affero GPL will ensure SaaS customers will have to buy something from them as well).

Finally, IBM just woke up with WebSphere CTO saying CEP is SOA next big thing. Once again few notes here... What Jerry Cuomo says is actually ...I really believe it's the next big thing in SOA, and that's event processing. So he jumps on the CEP wagon but in fact extends its scope to event processing to make the point. Well, who would disagree with that then?
I actually talked about the CEP - SOA - Event Processing relationship in my JavaOne Esper presentation back early 2007.
I am really looking forward to see what IBM will really deliver in that space, beyond research, research, research, forward looking announcements like SystemS or tiny limited features additions here and there (see ObjectGrid CEP f.e.).
I bet the next CTO will finally acknowledge that XTP meets SOA precisely through CEP and this one is really the next big thing - ie CEP is far from being just in SOA.
For some reason Jerry forgot to add XTP to his mix ;-)

2008 is going to be an interesting transition year for the CEP space as everyone is basically working on it growth, be it with open source, evangelizing, use case solving, or SOA bandwagon.

No comments: