Eclipse is dead. That's why, at a time when most conference organizers are, in Mike's words, "either canceling their events or losing their shirts", EclipseCon 2009 will likely turn a profit. That's why companies like Microsoft, for example, sent more attendees to EclipseCon than ever before, and Sun Microsystems (or is that IBM?) was actually a sponsor. That's why the bird's nest attracted more than 250 occupants (many of whom, like me, were first-time tweeters) in just a few days.
Modeling is dead, too. That's why I saw so many fresh faces at my EMF tutorial last Monday. That's why Peter's (Ed's) "stupid modeling" talk was so well-attended (and well-received, from what I've heard). That's why there were not only one, but two, modeling BoFs last week ("the sequel" and "the sequel to the sequel").
Yeah, Eclipse and modeling are "dead". You can go ahead and quote me on that. But if you do, please don't take my words out of context.
(Oh, and if you can't sense the sarcasm in this post, I'd be concerned that Eclipse and modeling aren't the only things that are dead...)