I am a little late to respond to this but Dare has a post that reiterates why web services should not be consumed in an RPC-like manner with tight-coupling to WSDL (even if you haven’t bought into SSDL yet).
A consumer should be able to make the decision to ignore messages it doesn’t understand without some auto deserialization process messing everything up. That is why web service consumers need to work directly with raw messages (XML or the actual SOAP message in this case) and avoid any of the proxy stuff that frameworks try to provide.
Once the consumer receives a message, the consumer can inspect the message and if it is not a message the consumer understands, it can simply discard it.
Keep using proxies at your own risk!
I’m attending Dr.Dobb’s Architecture & Design World Conference. Here are my notes from sessions I attended today:
- Interaction Testing – TDD is ultimately supposed to influence the design of classes. Code-first or code-after is not as important as letting tests lead to eventual refactoring. Code-first can reduce the amount of refactoring that may be required. Ultimately to effectively unit test, there will be lots of classes and interfaces with liberal usage of mocking for interaction-based testing. Only “state manager” classes will leverage state-based testing heavily.
- Creating DSMs – Domain Specific Models have the potential to significantly increase developer productivity from a “language perspective”. Focus on creating a narrow DSM with an appropriate code generator. There will always be code outside the model, this is the domain framework. Start small and evaluate the benefits of the DSM.
- User Experience and Interface Architecture – Architects do not focus on the “interaction architecture” of the application. Avoid inside-out design rather focus on outside-in design. Don’t try to outsmart the user, rather focus on the activities that the user will undertake and make those activities adaptive by the user.
Two days ago, the Flying Eagles of Nigeria crashed out of the U20 WC currently being played in Canada by the margin of 4 goals. All 4 goals were scored in extra time. Simply put, the team collapsed as extra time was being played. A couple of observations I have about the match I have include:
- Chile was fundamentally better, both technically and tactically
- The Flying Eagles lacked vision and purpose around the 18 yard box
- The Flying Eagles technical staff did not make timely and key substitutions neither did they change tactics
- We didn’t seem to have a “difference maker” on the team
Now these observations are not eye-opening in of themselves, but it calls to question the future of football in Nigeria (both players and technical staff). The Super Eagles are recycling the same set of players (it seems like) under the guise that they are “oldies but goodies” with very little infusion of youth.
Amazing what a blender can do!