David Luckham’s complexevents.com pointed me to a Dr Dobbs journal article by Adrian Bridgwater titled “Complex Event Processing: IT Liberator or Over-Engineering Hell?”, which starts with the question:
“[do] Complex Event Processing (CEP) tools and engines spells good news or bodes ill for our technology future. Specifically, I’d like to question whether CEP represents a highly agile customer-influenced IT delivery driver, or whether CEP will pervade (and ultimately “invade”) our use of technology to the point where it becomes insidious.”
The 2 options Adrian mentions are not mutually exclusive though: CEP and CEP-based information technologies can BOTH provide agile IT delivery AND be pervasive – although why that should equal insidious I don’t know. In some respects “event processing” is already pervasive – on your GUI’s, smartphones etc. CEP is in some respects just moving the advantages of the “event viewpoint” in IT down the chain towards other IT services.
Adrian goes on to mention:
“CEP technologies may soon migrate downward from mission-critical high-end deployments to affect the software that (for example) runs in the geolocation device in your cellphone… As you stroll through the mall, your device starts to alert you of special offers relating your consumer spend behavior…”
“May soon” is out of date and should be replaced by “Are already”! The “mall-offer-smartphone” use case is already done for at least one retail customer in TIBCO (and another such system was demo’d at TUCON in May this year).
The third and last problem with the article is that it mentions some small vendor’s hype – “programmers have struggled to tap into the fabric of events happening across the enterprise. This has left CEP engines lacking the data required to make an impact…” – which is such rubbish I’m amazed Dr Dobbs published it. Haven’t they heard of middleware? That pretty much all CEP engines plug into said middleware? And that such middleware effectively *embodies* the “event cloud” in many businesses?
Then I realised this article was in the Windows/.NET section of Dr Dobbs. OK, that might explain the thoughts above, but perhaps we would have expected at least one reference to the latest Microsoft offering on stream processing…
