Middleware Matters

I finally put my latest Internet Computing column (March/April) up on my home page. Written a couple months ago, this one is called "Web Services References," and it analyzes some of the issues with the endpoint reference as it was defined in the original WS-Addressing W3C submission. Fortunately, these issues have been largely addressed with the recent addition of the metadata abstract property to the endpoint reference (see the working group's last call working draft for details).

On a related note, the prolific and always insightful Bill de hÓra recently asked WS proponents to explain, among other things, why protocol neutrality is a good idea. Not only does the column mentioned above relate to that topic, but I've written a number of blog entries, papers, and columns over the years, such as this one, that explain fairly clearly why protocol-independent applications are desirable. It seems so obvious that I would in fact turn the question around: why would anyone explicitly choose to tie their business logic directly to a specific networking protocol?

The call for papers for the International Symposium on Distributed Objects and Applications (DOA) has just been posted. I co-chaired DOA the past two years, but this year I'm a PC member, since having the same chair over and over would tend to limit the conference's evolution.

Not only does DOA draw some interesting papers in its own right, but it runs concurrently with two other conferences to form the On The Move (OTM) Federated Conferences. This makes for some interesting interactions between groups that might never talk to each other. It's also typically held in some nice places, and this year is no exception as the conference returns to Cyprus, same place it was held last year. Beautiful weather!

I religiously read ACM TechNews because it always covers a variety of interesting topics, and the article summaries it provides are excellent. I was surprised when I was reading along and found that Wednesday's edition contained a pointer to my most recent Internet Computing column. Cool! My next column, which should appear around the middle of May, contains the second half of the coverage of web service references.

I also recently published an article in the Business Integration Journal entitled "What SOA Isn't: Debunking the Myths of Service Orientation" (PDF). It has a little fun at the expense of the "SOA" acronym, but overall it tries to describe what SOA is by explaining some common misconceptions that people have about it. Unfortunately, though, I had to leave out my favorite "not SOA" acronym, which is "Special Object Annotations," due mainly to space considerations.