Tim Bray and others have questioned the plethora of WS-* specifications.  Tim says:

I still think the WS-* stack is bloated, opaque, and insanely complex.  I think it’s going to be hard to understand, hard to implement, hard to interoperate, and hard to secure.

and then states how he wants to be the loyal opposition to the work going on in this space.

Are there too many specs?  If you think so, then there are.  So don't use them all.  You don't have to use any of them.  But if you need the capabilities they provide, like message-level security, then use the ones you need (in this case WS-Security).  The composability aspect of the Web service specifications is such that you can pick and choose what capabilities you need.  You don't have to use every spec that comes out.

Is the stack opaque?  I'm not sure I understand exactly what Tim means here, but I don't know how Microsoft and our partners can be more open about how we are creating these specs.  We are doing things differently in that we aren't going to a standards body first, then coming up with a draft of a spec.  This takes too long.  However, through the Workshop process we are getting public feedback, posting materials and results, creating discussion groups and bringing a well-fleshed out spec to the standards body for expedient ratification.  If you have an opinion on a spec, there is no reason why your voice cannot be heard.

Is the stack insanely complex?  The release of WS-Transfer and WS-Enumeration has understandably made a few  heads swim -- "Boy, more specs!  What are they thinking?"  But I cannot imagine that anyone would really look at either of these two specs and say that they are too complex.  They are trivial.  When I consider the work we did to make the initial SOAP implementations when it was a young spec and dozens of people stepped up to play around and create an implementation, I think it is safe to say that it would be significantly easier to create a WS-Transfer or WS-Enumeration implementation given todays tools.  Admitadly, they are specs, so they aren't something I'll give my wife to read for a little light-reading before bed, but they are pretty simple.  The same goes for most of the recent specs.  In fact, its probably safe to say that the most complex spec is WS-Security, which needs a certain level of complexity to handle the important issues it deals with.  But it just so happens that WS-Security was the first WS-* spec to be ratified by a standards body.

Is the stack hard to understand, implement and interoperate? Here's an article that explains the specs, but again, if you don't understand all the specs, don't worry about it.  Tools are being created by people everywhere to make it so you can just indicate the capabilities you need and the rest will be done for you.  If you look at all the major players signed on as authors on the specs, you get a pretty complete picture for all the popular platforms.  Add to that the people creating implementations that don't happen to be authors on the specs and you will find a lot of people dedicated to creating implementations.  And by and large, the specs are pretty darned simple.  A major portion of the standardization process (both the Workshops and the standards bodies' ratification process) is how well implementations interoperate.  If during the Workshop process implementations don't interop for some reason, the authors can go back and modify the spec to address any problems.

Will the stack be hard to secure?  I think the answer to this question will be answered very soon.  WS-Security has been ratified for a few months and implementations are being developed in a lot of different places.  Is it hard to make WS-Security interoperate?  This is probably the hardest of all the specs.  But there has been success in the past with early versions, and I suspect a lot of success as supported versions start being released by all the major vendors within weeks.  Obviously it is hard to look good if your WS-Security implementation does not interoperate, and there are definitely efforts being made by all the major players to test against the other implementations.  If 6 months from now we don't have implementations from most major platform providers that interoperate, I will be stunned and will question the future of the WS-* stack.  But I am really, really, really confident that this will not be the case.

   -Matt