Wednesday, October 29, 2008

On EDA, CEP and disruptive technology

Text Color
This is the second time in the last few months that the term "disruptive technology" is being used, this time by Mark Palmer who adds his voice to the EDA vs. CEP discussion. Mark acknowledges the fact that EDA and CEP are not synonyms, but asserting that CEP is the disruptive part of EDA.
Recall that in the Oracle Blog the "disruptive" word was used recently and I have discussed it at that time; Richard Veryard has answered my response claiming that "disruptive" is often not better, at least not in the short run.

Anyway, I have noticed in the recent few weeks, two IBM customers who made plans to move to EDA (neither will use CEP, at least not in the short run). The shift to EDA is a fundamental change in the architecture thinking with the introduction of the decoupled event-driven thinking. Is it new ? - not really. Is it new for these customers - yes, not only new, but a significant change in thinking, it is an indication that there is a beef in EDA, even without introducing CEP, that the enterprise should digest. While thinking in events is natural in the daily life, it is still not natural for enterprise architecture and programming paradigm.

Bottom line -- while CEP certainly has its merits (if the customer has mature enough to digest it), EDA seems to be a more fundamental change in thinking relative to the alternative, and has its own beef. More - Later.


Mark Palmer, StreamBase CEO said...

Great post, as usual Opher. And I love the reference to the "beef" of EDA. Agree that there's a lot of beef in EDA. In fact, personally, I think I'm more inclined to be dismissive about EDA because I started my professional experience with software in 1988 building trading systems. These systems were EDA before we called it EDA, so it was my baseline for comparison. In fact, it wasn't until later that I was OK with people who wanted to an RDBMS in the middle of a software system - they just slowed things down!

But that said, you, and the Jack and Joe are right that EDA is a big jump for some people, and that, as I readily offer, CEP is not a pre-requisite to a great EDA; indeed, my own argument makes that point: we had excellent EDA for 15 years in capital markets without CEP!

So I'm happy for this debate because perhaps together we'll raise the visibility of EDA. And once you get serious about EDA, even in a basic implementation, I think the architectural journey will naturally lead you to CEP.

Anonymous said...

Hi Opher,

Good comments. I agree, for the most part, here:

Yours faithfully, Tim