Box, an architect at Microsoft, urged developers to focus on innovations at the application level.
"Unless you're a big platform vendor, the place where you should be thinking about innovating is applications.
"The plumbing is boring" and is nearly completed, Box said. He cited the recently announced Google Web APIs service, for conducting searches over Google from within applications, as an example of an exciting application.
Network infrastructure also presents an opportunity, but the basic Soap stack itself is done and another one is not needed, Box said. The tools and plumbing are already good enough for "killer apps", he said.
Soap, he said, arose out of an HTTP foundation, with HTTP being a protocol accepted by multiple development camps. XML also was critical.
"Really, the novelty of Soap was simply saying we're going to take XML and make it work for data," Box said.
An upcoming version of Soap, Version 1.2, needs to be the final version, Box said.
Box also stressed that XML Schema is the dominant technology for Web services. "Had XML Schema been done in 1998, we would not have done Soap," he said.
"The reality is, XML Schema is the foundation for the rest of XML," said Box. Technologies such as XML Query take XML Schema "for granted", he said.
Soap messages need to function with XML, according to Box, because XML Schema is an "inevitability".
"There is no point in not embracing this thing and I strongly encourage those of you who work in Web services technologies [to] make sure your story is straight," with respect to XML Schema, Box added.
Box also addressed some critics of the UDDI (Universal Description, Discovery, and Integration) specification, saying UDDI can growth and that the upcoming Version 1.3 of UDDI hopefully will fix problems with identifiers.
"It's hard to talk about UDDI without taking an apologetic tone," Box said.