Ideally, we should base standards on implementation experiences, and users—not vendors—should write them. When a vendor writes a standard, the product that the vendor eventually wants to build and sell is often what becomes standardized. Sometimes, they even wait to build such products until the standardization is nearly complete, to avoid wasting effort building something that will get changed during standardization. The end result can be standards that are full of holes and ambiguities due to the lack of actual implementation experience.
As usual with Steve’s writing, the article is pretty great and definitely worth reading. I basically agree with everything he says in this particular case — but I believe that regardless of the validity of his arguments, there is simply no way that Microsoft, IBM and BEA together could fail to dominate the ‘standards market’. Their combined power is simply too big, and if they agree to introduce WS-JumpOffACliff, the industry will gladly accept it. Smaller vendors, such as IONA, probably stand a better chance of influencing the specs by participating in the feedback workshops, even though clearly an open process would be more democratic.
However correct Steve maybe in asserting that the users should be the ones writing the standards, I clearly can’t see them doing so, at least not in a significant manner — after all, what would be their incentive? The industry’s greater good? Come on.