A few of us have been discussing event processing languages (EPLs) for a number of years, advocating that SQL-like languages are appropriate for certain classes of CEP/EP problems, but not all.
Some readers might recall that I published a draft survey on EPLs to the Yahoo! CEP Interest group titled, (DRAFT) A Survey of Event Processing Languages (EPLs), October 15, 2006 (version 14).
A number of us CEP “grey beards” have consistently advocated that there are EPLs and analytics that are optimal for certain classes of event processing problems (and, in turn, there also are EPLs that are suboptimal for certain classes of event processing problems).
For readers who do not frequent the Yahoo! CEP group, below here is a link to a copy of the original survey.