All of these companies are piling into this real weird bandwagon. ESB is a problematic term that was sickly in 2005 and I think it's going to die in 2006.

SOA is maturing. If you wanted to do SOA four years ago you were pretty much on your own. You had a couple of products that would let you do some Web services, but now SOA is a lot more than Web services. It's governance and security and metadata and much more.

This is just continuing the momentum. We are past the stage where SOA is hype. It's the right time for SOA. Now it's a matter whether or not it will stick and whether companies will commit to it. Success stories like this help.

That's a bit less sexy, because you're getting down to brass tacks: implementation details, not the big news stories.

It's another indication of the market maturing. Clustered security is pretty advanced.

It's not so much sizzle now. There's more meat.

There's always a context to data. Even when a field is blank, different applications impose different assumptions about what that means.

The way we think about interacting over the Web is changing. It only took one application to alter our entire expectation level.

The whole portal approach to loading lots of stuff on back-end Web servers has grown stale in the tooth.

Do you divvy it up, or do you provide a central service? SOA raises the fact that data is heterogeneous.