Our industry WS-* leaders, and all of us (Web and Web Services advocates together), should have a good look at this poster and perhaps take a moment to reflect… where did all go wrong? Why are there specs doing the same thing in a slightly different way? Why some of the specs are so unnecessarily complicated?
This should have been easy!
6 responses to “The WS-* specifications”
Part of the problem is that there’s no single body overseeing the “architecture”. There’s no equivalent of IETF or OMG. W3C and OASIS are fine, but only W3C has a notion of a Web Services architecture (which is now horribly out-of-date anyway).
You guys have forgotten the KISS principle.
Some spec duplication is simply a marketing/”me too” issue on the part of some companies.
Security, reliabiilty and transactions are difficult problems in any system. Now make that system distributed and loosely coupled – the difficulty level raises substantially.
The key word to remember here is composability. Use what you need for the solution at hand and ignore the rest.
It’s a mess not because of lack of oversight, but because of the lack of use of principled design techniques.
Mark, I think your comment “because of the lack of use of principled design techniques” is a bit sweeping and overstating things. I obviously can’t speak for all of the WS-* specifications, but I’m confident this doesn’t apply to the ones I’ve been involved with. Those ones really do have more of a political agenda around them. Having done this kind of work through the OMG for over a decade, I’m pretty confident that an organisational structure like that would have been of benefit to the community.
Actually I think I just made a sweeping statement too 😉 Some of the specs. I’m on (like WS-Addressing) do fall into your category. But I still believe that a structure like that within the OMG would help to impose a better order on the chaos.