In a meeting today with a lot of IBM Websphere product architects they didn’t quite come to the conclusion that UDDI was dead, but they did say that at this point UDDI is an important spec, but it’s just not meeting the needs of people trying to roll out SOA architectures.
IBM’s thoughts around this assume that everyone is coming to the same conclusion that UDDI just doesn’t do enough so they’re creating more functional replacements. The replacement is called Websphere Service Registry and Repository Capability at this point. I’m not sure whether there are more WS* specs or Websphere branded products in the world.
At this point they didn’t have a whole lot of details around there product in particular, but they were willing to speculate on where UDDI is going. There idea is that a few proprietary approaches are being developed by vendors and then a spec process will happen after some experimentation and everyone can essentially agree on the UDDI replacement or UDDI 2.0.
They didn’t really seem too out in left field on this as our company architect has been telling me UDDI is pretty much dead on arrival for a while now. In addition at a Bird of a Feather session on SOA at SD West 2006 I asked how many of the 50 participants were using UDDI. One brave soul raised his hand from the State of Oregon. He said they did have a UDDI registry setup and a few web services registered with it. He then explained that everyone using their web services had hard coded them without using the UDDI registry at all.