WSDL Debate

I've been keeping a half, or perhaps a quarter, of an eye on this
seemingly interminable and increasingly acrimonious debate about
Semantic Web Services and WSDL.  Here are some random thoughts about
this subject, which I post with some hesitation because I question
wether this discussion is really getting anywhere useful.

It seems to me that one should look to commercial applications as the
primary user community for Semantic Web Services.  Businesses.  Although
I understand that there are some legitimate use cases for REST style Web
services, in most circumstances I think that we are talking about the
WS-* stack of specifications that depend on the SOAP headers.  I also
think that at least initially such use case are likely to be on
intranets, not B2B, simply for practical reasons.  There's plenty of
potential value on corporate intranets.  

There have been tremendous resources put into developing the
constellation of WS-* and WSDL specifications and implementations
associated with the current Web services stack.  I don't think that it
is reasonable to expect current implementors to throw all that stuff
away and replace it with something completely different, and the cost of
duplicating all the necessary functionality in a new platform would
itself be prohibitive.  I think that whatever Semantic Web technology
that is applied to Web services MUST be implemented as an add-on to what
is already there.

It seems to me that it would be REALLY good to generate, as quickly as
possible, some incremental and possibly modest, but nonetheless useful
semantic additions to the current WSDL situation that can actually be
deployed immediately.  Hopefully such a first step would be done in such
a way that it would be extensible to more sophisticated scenarios, but
the key, in my mind, is to GET ON WITH IT.  There are a whole bunch of
reasons why I think this, but just to focus on one of them -- I assume
it has not escaped your notice that Microsoft is not playing in this
Semantic Web space at all.  What does this mean?  I'm not really sure,
but one thing it means to me is that I think you don't have five years
to come up with something deployable.  I know that Microsoft is quite
capable of making abrupt U-turns and picking up on things in a big way,
but I don't think that they are very likely to do so unless it's
something with quantifiable value that can be deployed immediately.  Why
should you care?  Well, I suspect that if Microsoft does NOT make such a
U-turn that they are perfectly capable of putting something into the
market within the next few years that will seriously blindside the
Semantic Web efforts.

Received on Thursday, 16 March 2006 22:43:56 UTC