Two ex-teammates, Tim Mallalieu and Jeromy Carriere, have a new article on Enterprise Interoperability between .NET and J2EE posted on MSDN Architecture Center. Combined with Simon’s book and the Patterns and Practices interop guide, there’s alot of good guidance on this topic.

My team owns the architect track for TechEd 2004. We’re hoping to build on the very successful Architecture Symposium from PDC. As part of that effort, I own that community efforts for the architect track. I’m working on the plan now. Obviously, we’ll have “the usual” events: Ask the Experts, Birds of a Feather, Speaker Lounge, Attendee Party, etc. Any other suggestions? Please email me or leave a comment.

First Simon, then Ram, now another member of the .NET Enterprise Architecture Team is blogging. David Hill just started, but he’s already posted an article entitled “A Simplified Asynchronous Call Pattern for WinForm Applications“. In the article, David describes “how you can implement a simpler asynchronous call pattern which allows you to consume web services from a WinForm application without having to worry about threads ever again.” Cool!

Another member of the .NET Enterprise Architecture Team, Ramkumar Kothandaraman (we just call him Ram), has started a blog. Like Simon, Ram spends most of his time with customers solving hard architecture issues. Check out his post on architectural agility.

Michael Earls is pleading with MS bloggers to focus on released bits instead of all the cool future stuff we previewed @ PDC. He is especially frustrated with my response to Scoble’s post about how syndication will look in the Longhorn timeframe. Mike, I can’t speak for Scoble or any of the other MS bloggers, but I’m sorry that it’s been hard to keep up. You’re a member of my target audience, so it’s good to know where we are missing the mark.

In my post, I said that Scoble shouldn’t focus on how syndication evolves in the Longhorn timeframe, rather how it evolves in the face of Service Oriented Architecture. And as watered down and nebulous as the term SOA is, when I use it I’m not implying that you have to wait for Longhorn. Indigo will be a great platform for services, but that doesn’t mean you can’t do them today. In fact, the advice coming out of my group is to start doing services right away. Check out the first two sessions from the PDC Architecture Symposium (here and here). In 110 slides there are only about three slides that mention Longhorn, Indigo or Yukon. The rest of the slides are focused on “Practical Advice for Building Your Services Now” (PPT deck from first session, slide 5). Stuff that you gotta worry about regardless of the infrastructure you’re building on. Tentative Operations. Avoiding Ambiguity in Messages. Stability of Data and MetaData. Service Masters and Service Agents. Mike, when you get a chance, please check out those sessions from the PDC Architecture Symposium and then let me know if they can help you right now.

It’s too bad the chapter on SOA that’s posted on MSDN is from a book on Longhorn. That implies you can’t do one without the other. Truth is, you can build traditional tightly coupled apps with Longhorn and, more importantly, you can build services without Longhorn.

What’s weird is that I’m actually not dogfooding anything right now. Oh, I have a VPC with the PDC Longhorn bits and another with Yukon and Whidbey installed, but I haven’t spent much time with them recently. I guess I didn’t make it clear in my recent post on SQL Service Broker that I can’t wait for it because I’m not actually using it yet. The only beta software I’m running on my host machine is Firebird and Thunderbird. I have two primary development VPC’s – XP and WS03 – and the only beta stuff running in either is WSE 2. Call me a slacker, but Whidbey/Yukon/Longhorn aren’t far enough along yet for my current projects.