I want to shed some light on why the STK supports DIME and not SwA. This topic has come up on soapbuilders. The purpose of this post isn't to defend the party line nor to dispute it. It is just an attempt to explain how we ended up where we did.

Many people may not know that the STK is a product of the Webdata XML team at Microsoft. We do System.Xml, MSXML, and various other pieces of XML infrastructure. A different team is responsible for Microsoft's SOAP and XML Web Services vision and the implementation of that vision in the .Net Framework. How the Webdata XML team ended up doing the STK is an accident of history.

The STK is a tactical product. It was concieved as a solid standards compliant and interoperable SOAP implementation to replace the SOAP Toolkit 1.0. STK 1.0 was an MSDN sample SOAP implemenation that wasn't very standards compliant. The .Net Framework is Microsoft's strategic platform for SOAP and XML Web Services.

The STK turned out to be a bit more useful then some people thought it would be. A number of customers used it for real life applications. The STK team got excited about it and wanted to improve on it. One of the more common customer requests was for better performance when sending blobs and large XML documents. The obvious solution was to add support for SwA.

While we were working on SwA, the web services team decided that DIME was the strategic direction for encapsulation. DIME is a better technical solution. It can be implemented much more efficiently and with simpler code then SwA. They also wanted the WSDL Extension for SOAP in DIME to be simpler then WSDL's MIME binding. There is also the potential for confusion between the infrastructure that can be provided by MIME and the infrastructure that will be provided by GXA. DIME solves one simple problem while MIME solves many problems, many of which are addressed in more appropriate (XMLish) ways by GXA.

Our customers needed SwA for interop, but if we released support for it in the STK, it may have made SwA a permanent part of the web services landscape. The STK team was forced to make some hard decisions. After much discussion, we decided to support the web services team's long term strategic direction. We replaced the MIME support we had partially completed with DIME support.

Was providing only DIME support in the STK the right thing to do for web services in the long term? I hope so. Was it the right thing to do for the STK's customers in the short term? No. I'm sorry we could not provide the SwA support they needed.