+1 XML enjoyed great success not because it has more functionalities than SGML, but it was easier to implement and use. I think Clayton M. Christensen's book "The Innovator's Dilemma" was mentioned on this list a couple of few weeks back, but the example of the hard disk market in the first chapter reenforces this thought. Smaller disk drives didn't offer larger disk capacities, lower cost per megabyte, and faster access time; it offered something different - smaller physical size to fit into smaller personal computers.
Likewise, I think BPEL is something to build on - I don't think it's suppose to be an all purpose process scripting language for all occassions. Similarly, I know that some vendors are trying to make their ESB products into a "all purpose" product, but I view it just as a base platform to connect different technologies together. Is SOA about replacing the current technologies with a new one? I thought it was about extending the use of existing technologies while offering use of new technologies. As the hard disk market wasn't about larger disk capacities, lower cost, faster access time, I don't think the SOA market is about offering "better" technological functionalities then the current technologies. H.Ozawa Sanjiva Weerawarana wrote: > Similarly, BPEL was designed as a base with the idea that vendor groups > would define extensions (activities, scripting languages etc.) which > would suit such group's purposes. Have those happened yet? I think Frank > pointed to a few starts but AFAIK things are not prime time yet. I don't > think you should throw the baby out with the bathwater. > > Sanjiva. >
