Walking up at 3:30 am to get to MIA to catch a ride to dallas. Urgh!
Missed the initial keynote, but managed to grab Don's early
talk, which was quite fun, as Don's talks always are. One interesting thing mentioned
in the talk is WSE 2.0 getting a new programming model allowing explicit access to
the SOAP messages. Will look forward to that one...
I'm sitting right now here listening to Doug
Purdy talk about serialization and loose coupling. One interesting comment Doug
brought up is "To achieve loose coupling, you're gonna need to understand XSD". True,
right now at least. But I have to wonder about this for a little bit. Do we actually
need to fully understand XSD? It's pretty complex, and some of the XSD constructs
are pretty obscure. One has to wonder if we only need to understand XSD good enough...
Good as in an XSD subset that maximizes the benefits and minimizes the cost, which
for all practical terms means getting the significant XSD subset that both allows
a decent level of flexibility while easing interoperability. I'm not one to usually
run away from complexity, but I'm not one to freely embrace it unless I see a real
need for it....
