SOA is in a similar state. I am actually very happy that SOA is dead (or at least, it is acknowledged as dead by many of the commentators). It shows that SOA is now moving into serious prime time. The hype is over, only somber reflection and execution remain. Move past the TLA, please!
In a prior job, I went through a similar cycle while evangelizing SOA. Predictably, I got nowhere. But my team and I went ahead and built an application very quickly by leveraging pre-built services. (Ok, you SOA purists, I know SOA is more than that, but we are talking about 2004 here - cut me some slack, will ya?) When my boss, the GM, surprised at how quickly we deployed new functionality, wanted to know the secret of our speed, I told him, "Remember SOA that I kept harping on? Well, that's what we used."
I was in southern Sweden last month, talking to a SOA team about SOA adoption. Among other things, they wanted to know how to "sell" SOA internally. The best advice I could give was, "Stop using the word SOA."
Whoever thinks SOA is really and truly dead doesn't quite understand what it is all about. There will always be a king, but we may stick different labels on him (Czar, Party, Messiah, Emperor, President, etc.). It is actually very good news when one of the biggest roadblocks to adoption is the label itself. (I can't help direct you to "SOA Adoption for Dummies" for a very sensible and practical treatment.)
The only odd thing Anne's post is the sentence: "SOA is survived by its offspring: ..., BPM, ..." BPM an offspring of SOA? Someone has to explain that one to me.
Now I am just waiting for BPM to die.