Web Things, by Mark Baker

Hey look, a WS-spec built largely upon a state transfer based architecture. Who knew? *groan*
(link) [del.icio.us/distobj]

Ouch! 8-) “As far as the web is concerned, the WS-* work is about sprinkling XML pixie dust on a failing idea.”
(link) [del.icio.us/distobj]

It’s not all bad news apparently. While one activity of the W3C treats N^2 integration complexity as a selling point, a new one fights hard against it. Talk about split-brained!

IMO, before any new WG is chartered at the W3C, the objective of that WG should be validated for architectural consistency with the Web (you know, the thing it’s trying to lead to its full potential?).

Here we have a WG whose primary objective is to produce a spec which will, prima facie based upon its charter, violate a key architecture principle (what, it’s just “good practice”?) of the Web; identify things with URIs. How can you get any more architecturally inconsistent than that?

I’m not upset, I’m just confused. How has it gone this far? Mass hysteria should not be an excuse. It’s a good thing that OASIS seems to have done the bulk of of the WS-* heavy lifting, leaving just four at the W3C. But that’s three (SOAP is goodness) too many, and it could have been much worse for the W3C.

Anyhow, best of luck Mark, you’re gonna need it.

Update: just noticed this comment from Jorgen;

As they say, “You can’t finish unless you start”.

Unless you’re finished before you start, of course.