Monthly Archives: February 2004

WS-CAF Presentation from Edge East

Here’s a link to the WS-CAF presentation that I gave at Edge East Feb. 24 in Boston.
Edge East 2004
You can find another, more detailed one here:
WS-CAF in a Nutshell by Mark Little


Hello from WS-CAF

Welcome to my new blog!
I’d like to thank Steve Vinoski for helping to set this up for us. If you haven’t read his blog, I highly recommend it.
I’m writing this from the face to face meeting of the OASIS Web Services Composite Application Framework (WS-CAF) technical committee, hosted by Attachmate in Paris.
Paris is cold and rainy, with a chance of snow. Which is unfortunate, since my mother is with me. I promised to bring her to Europe some time, and because I know she likes Paris, I thought it would be a good idea. Oh well, I’m sure we’ll have a good time, even if it is a bit cold to be out walking around.
I got here a little late because I gave a presentation about WS-CAF at Edge East on Tuesday in Boston. I’ll post the presentation to the site as soon as I can figure out how to do it ;-). About 25 people attended, and they seemed pretty interested.
WS-CAF is a standardization effort focused on what I consider to be one of the last major unresolved areas for Web services – transactions support. There have been several attempts so far, but nothing has yet really cracked the problem.
Just after SOAP 1.0 came out, Don Box and I tried to map the Transaction Internet Protocol (TIP) to SOAP, but quickly discovered we had bigger problems than just putting the two specs together – namely the lack of persistent sessions in HTTP. TIP, like any other current distributed transaction protocol, depends upon persistent sessions. This is one of the reasons you often here statements like “two-phase commit doesn’t work well for Web services.”
Also, a Web service can interface to any technology – application servers, messaging engines, packaged applications, databases, scripts, etc. – and these technologies often have widely disparate transaction models and protocols.
I think we finally may have the basis for the solution to both problems.
In WS-CAF is the WS-Context spec solves the former problem, and the Business Process (BP) transaction model in the WS-Transaction Management spec solves the latter.
We’ve worked hard to ensure that WS-CAF is a compatible superset of WS-T, WS-C, and WS-BA released by Microsoft, IBM, and BEA, since those are important specifications, too.
During an interview at the Edge East show, Jeremy Geelan said he noticed some passion in my voice when I talked about standards.
It’s true – I’ve been working on various software standards for longer than I like to think about – fifteen years or so – and I would still really like to see the software industry standardized, as other industries such as manufacturing and shipping have, so we can all realize the benefit of “mass produced” software. More later…