UDDI will be the building block that enables companies to quickly, easily and dynamically find and transact business with one another, according to its backers, which include Ariba, IBM, Microsoft and about 200 other software companies. The problem is that in many cases, organizations will be reluctant to use this new functionality, especially in the brick-and-mortar world.
Read UDDI's Problem: Technology Cannot Replace Relationships
There are two capabilities that web services model (based on ebXML or WSDL/UDDI/SOAP technologies) promise to provide B2B with. One is to enable Just-in-time B2B integration (which is still just a promise, especially on WSDL/UDDI/SOAP side) and another is to reduce the TCO factor for B2B integration.
Although I agree with the author that JIT B2B integration would be just a promise and that manufacturers and suppliers/retailers would hesitate to dynamically find trading partners and start doing business with them right away, I think businesses should and probably would take benefit of Web Services model in B2B for reducing TCO factor of integration. Just imagine, a supplier (belonging to SMI category) who had been involved in business collaborations with trading partners using email/fax/telephone/fedex (because they would not be able to afford EDI), would now be able to do B2B "on wire" with these trading partners in a more sophisticated way (and affordable too!). I think this is where B2B would require these web services technologies.
Just my view of the world...
Let's not get over-excited about dynamic web services and think that technological capabilities should rule over common sense. Just because you can do something with technology doesn't mean that you should or have to. I don't know any business that will switch suppliers on a whim just because UDDI and WSDL MIGHT give you that possibility. Web Services reduces the cost and eases B2B integration. Past that ..well someone is going to have to show me what is next. Video-on-demand, WAP,.. the techno world is littered with what marketing people thought were good ideas at the time..
I agree with Eddie. Technology cannot drive functionality. UDDI will probably serve as a good directory of web services available with a prospective partner, categorized in a predictable manner - a more organized version of today's web search engines. Next, its upto the business guys to check whether the prospect is good enough for a partnership, just as it is in today's world. And then would come the intergration of the partner's web services into the application.
Glad to see that I'm not the only one not chanting and preaching the take over of the UDDI/WSDL paradigm. Notice I left out soap in that combo as it's simply a message envelop and a practically solution for communication via the http (and other transports!). Soap, ebXML, and cXML I see as being viable solutions for needed integrations, some technologies being more prevalent then others. As I posted in another news item and as you guys more gracefully state here, a business is not going to drop everything with established contracts and open up the great big yellow book in the sky! Moreover UDDI and WSDL was a means to integrate the million and one over night eCommerce shops popping up.
All it's gonna take is for the bean counters in these organisations to see the cost savings that can be made. Once that happens anyone who doens't use it is going to be left behind. Like any advance, it starts as a trickle and ends in a flood.
The author is bang on when he says that B2B buying/selling is too involved for JIT approaches that UUDI enables.But IMHO,the real key to using the UDDI/SOAP/WSDL framework is exposing critical biz intelligence hidden in back end systems to EXISTING trading partners.Imagine if Intel's manufacturing systems were to know Dell's inventory position in real time! That kinda real time biz intelligence will enable tight and intelligent coupling throught the value chain.
Also technologies like UDDI/WSDL/SOAP/XML will likely to help fulfill the promise of componentware,which has so far failed to really take off despite all the hype.
Another thing that technology can't do is get people to agree on interfaces. What use is registering your implementation of an interface in a discovery directory if yours is the only implementation of that interface? Anyone who new of the interface would already now where to find the implementation.
Service discovery is a cool idea, but it has never got off the ground before. Jini and UPnP are other viable service discovery technologies of which hardly anyone takes advantage. When Sun proposed Jini, they developed standard printer and file server interfaces, and invited 3rd parties to get organized and develop more interfaces. It seems that no one of any significance accepted their invitation.
It may not be for lack of trying either. Standardization isn't easy. There are many barriers, including attempts to protect trade secrets and NIH. Whatever the reason, consensus is rarely achieved. And without it, UDDI, Jini, UPnP, or any other incarnation is doomed to failure.
What we really need is a consensus technology.
UDDI = Software yellow page listings.
It's been a long held dream, but I point to the CORBA Trading service as a failed attempt at it.
Software services yellow pages will probably eventually take off in some form, though, since it's so appealing to have a software services registry for your perusal at component assembly time.... look at Yahoo.com
BUT overall, WHY a new protocol? UDDI isn't going to be enough to do "automated" federation of components, it will always be done by a human component assembler for the forseeable future... if this is the case, HTTP + HTML works quite well too, (again, see yahoo.com) and is probably less complicated.. what we need is a yahoo for web-services, not a bunch of protocols...
I am glad someone is bold enough to write an article against the current buzz. SOAP is a good idea. UDDI, on the other hand, is eerily similar in concept to BtoB exchanges and online auction houses (I am not referring to eBay). Connectivity and integration between software products is achieved via partnerships and not by looking up Yellow Page entries.
Some of the web services hype, especially UDDI is way out of proportion. I suffered through enough sitting in otherwise quality presentations at JavaOne, where Sun employees were forced to twist their presentations (in a very obvious and contrived manner) to include a slide on web services.