avasseur (1) [Avatar] Offline
It's great that you are including CEP in the book.
I have read the free chapter available from http://java.dzone.com/articles/interview-with-jeff-davis-auth and I would argue around a few updates. I am a long time committer on Esper and - Jeff - you can probably reach out to me by email at gmail if you want/need to.

1. The acronym that is more adequately terms for event processing is more CEP than ESP, where you seem to use more the ESP one. At least Oracle, Tibco, Progress, most of the academics and pure players like StreamBase or Coral8 are using CEP / Complex Event Processing in their docs. IBM and Tibco are also using Business Event Processing but it seems to me consensus is on CEP.
Refer to the Event Processing Technical Society EPTS proposed glossary

2. As of Esper 2.x the EQL is not used to describe the language as Event or Esper Query Language. Instead all the documentation is using EPL as an acronym for Event Processing Language - which is also a consensus among the eco system to talk about the language introduced to write the continuous queries (always proprietary in the sense that no standard exist and each solution comes with its own flavor)

3. For samples etc. I guess it is a bit too late but bear in mind that Esper 2.x introduced a package naming change. This should only be a refactor/rename on the client API for any Esper 1.x app.

4. I have noticed that you balance the commercial add-on approach that the sponsoring company EsperTech is using. To its defence, I believe MuleSource does provide non open source management modules, SpringSource does provide non open source productivity tools, JBoss did provided non open source JBossON prior RHEL (don't know the status there these days), and that this is a fairly common approach to ensure viability of a project in the long run.
See here for MuleSource, and see good generic analysis from Matt Asay