I do think that dynamic languages can help us tame the complexity of server-to-server communication on Web services networks. The reason: we simply don't know which arrangements will prove workable and which won't, and we'll need highly productive RAD tools to plow through lots of experiments. By the same token, we know very little about how to best enable people to interact with Web services. Here too, we'll need all the productivity and flexibility we can get.
This is something which is already causing me trouble with my Eiffel -- a decidedly static language -- implementation of the modelling framework I designed in the course of my PhD.
Comments