My notes show that I've been aware of OMeta since mid-'09. Today I went to read the original paper (DLS'07; top of above link) and was surprised to find I had never read it. I'm not sure why this is:
a) Because I first became aware of it as part of Alan Kay's VPRI project, but got lost in the details of COLA.
b) Because I had to refresh my memory on PEGs and then got sidetracked.
c) Because I ended up at http://www.bayfronttechnologies.com/mc_tutorial.html reading about meta II that OMeta is said to be based on. At least for me this turned out to be a bad idea. Several times in the past 2 years I've tried to make sense of that tutorial and failed to figure out the basic question: how is it different from a parser? Then again, YMMV.
The VPRI project over-emphasizes the past almost to the point of seeming to live in it. It's all very well to be aware of history, but beyond a point stop telling me that it's a reinvention of 1970s technology, just tell me what it is! I hung out for a few days on the FoNC mailing list (http://vpri.org/mailman/listinfo/fonc) and left when I remembered this complaint.
Anyhoo, after all these distractions when I did finally go to the source I found it surprisingly easy reading. The biggest hurdle was just figuring out what the heck it was, and how it fit into the concepts I was already aware of. The title of this thread summarizes my current best insight on that question.
Fun exercise for the bored reader: find the shortest previous sequence of links from the OMeta site to the arc forum.