Sean Rhody

It's December, and you know what that means: holiday wishes and New Year's predictions are due. That's right, once again we'll gaze into the WSJ mystic crystal ball (okay, so it's a Christmas ornament - we're on a budget here) and come up with our prognostications and pleadings. So, without further ado, here's the list.

This will be the year we finally get some traction on Business Process Management - you know, that thing everyone is talking about, but no one is doing yet. To do this, one or more of the solution providers is going to have to bite the bullet and do a decent job of implementing BPEL, along with WS-Transaction, WS-Choreography, and WS-Orchestration. It's time to stop hashing it all out and implement it - and make it interoperable.

One million more WS-* standards will be released. Most of them will overlap with each other. Few if any will be implemented. None will achieve the widespread acceptance of the basics of Web services - XML, SOAP, UDDI, and WSDL. While I'm exaggerating the number of specifications and standards that will be released, the reality is that there are already too many standards. Here's my holiday wish in this area: stop now. Don't create yet another standard, we have plenty. Work to extend the standards we have, as they cover most of what we want to do. I'm not discounting the significance of things like WS-Reliable Messaging, but really guys, come on. No one can keep up with more than a half dozen or so standards when it comes to Web services - we need one for discovery (UDDI), binding (WSDL), transaction management (legion), business process (also legion), security (legion as well), and maybe a handful of others. The rest should be rolled into these overarching categories. Why? Because then it would force vendors to actually implement some of this, instead of just creating new standards every couple of weeks. That would be good. I've said it before, but I'll say it again: we don't need more standards. So stop already, and implement - let the folks in IT catch up.

Service-oriented architecture will continue to be the buzzword for the year. SOA will be adopted by more organizations, including a number of those who should not go there. Service orientation offers the possibility of real advances in IT - advances that can be coupled strongly to business value. It also requires organizational change and modification to IT governance that many folks have yet to really tackle. It's a lot easier to change the software than it is to change the people who run the software - even when they're willing to change. It should be fairly intuitive, but at the risk of stating the obvious- you cannot buy an SOA. You can buy technology components that make SOA easier to implement, but you cannot buy the whole thing - it requires people, design, business involvement, governance, change management, and especially a measurable ROI. If the ROI is not there, don't do it.

Now, lets move on to my wish list. I wish Web services had a real, multivendor, language-neutral, XML-based user interface approach. No, sorry, AJAX is not it - it's not language neutral, it's not really about building user interfaces, it's more about asynchronous messaging, which while necessary, is only a piece of the puzzle. We really do need this - a way to design user interfaces for services, eliminating the bad parts of HTML while maintaining the download-only nature of the browser for ease of application deployment.

I'd also like to see some services. By this I mean I'd like to see OASIS or the W3C or some similar organization (RosettaNet would be a good candidate) start to define AND implement a set of business services, using Web services as the technology. Creating composite applications and really leveraging the power of an SOA would be a lot easier if standard services were defined. Notice I called for standard services, not a services standard - please, not another standard!

Last, I'd like a Jaguar - but I'll settle for you all having a safe and happy holiday season.