Questioning UDDI. What about WSIL?

Richard Monson-Haefel points out a well articulated piece "Who Cares About UDDI?" by David Chappell. Like Richard, Chappell does an admirable job of summarizing some of my misgivings with the UDDI specification. When I wrote my last paper on the practical uses of Web services the issues I was astounded that the UDDI was unilaterally supported as part of the "holy trinity."

UDDI seems much too grandiose and misses its true purposes. If you look at the timing and the originators of the initial spec you get the specification you get the suspicion that they may have still been drinking the B2B marketplace kool-aid. (Herein lies the problem of close specification design that is becoming all to common.)

I agree with Chappell that UDDI doesn't offer much value at present. However I disagree that a central directory services have little value until and if dynamic Web services become common place. At a minimum a directory will help large distributed organization manage, track, organize and maintain the presumably hundreds and perhaps thousands of Web services that will appear behind their firewall and with its close business partners. Given the IT hairballs I've witnessed this could be very valuable. Discovery is also another important aspect to these directories. It is not uncommon for one part of the organization to not know what another is doing that could be useful to them. If implemented and maintain properly I think the directory could resolve many of these disconnects. Does UDDI do any of this well in its current state? I think not.

I'm glad to see someone else noting WSIL (Web Services Inspection Language) as I did in my paper. WSIL, while developed as closed specification by the shadow specification organization call Microsoft and IBM, is more useful and requires less overhead to implement in achieving a similar results to UDDI. WSIL in a lot of ways is quiet RESTful in nature -- its is like RSS for Web services. For more of and in-depth comparison of WSIL and UDDI I recommend Tarak Modi's "WSIL: Do we need another Web Services Specification?"

I've been quite curious as to why I have not heard more noise about this seemingly useful specification. Maybe I'm missing something and need to dig further.