Two very different people -- Burton Group's Jamie Lewis and O'Reilly Media Inc's Tim O'Reilly -- took on similar themes during keynote addresses at the recently concluded Catalyst Conference (Lewis) and the Open Source Convention (O'Reilly). In essence, both said the computing stack had grown, Web services was at the top and everything else was on its way to commodity status.
Wayback when, the computing stack consisted of three parts: an application designed to solve a problem, an operating system and hardware required by the operating system. Over the years we could argue about various CPUs such as the Z-80, 8080 and 6502. But today, processors are commodities that better act like an Intel CPU. Then there were the operating system wars, going back to C/PM, DOS, DR-DOS and M/PM. The GUI saw its own battles.
But today, supposedly, abstraction layers and middleware are designed to look like a black box to the top-level Web services. We've even created the term service-oriented architecture to describe the commoditized stuff, the plumbing, if you will, that supports the services exposed to the user.
It's as if these services were the rock stars of computing and everything else had to knuckle under to their demands. It's true that Van Halen's standard performance contract contained a provision calling for the band to be provided with a bowl of M&Ms, but with all the brown candies removed. Not that David Lee Roth hated brown, but it was an easy way to quickly see if the show promoter had read the contract. So far there's nothing like this in Web services, as far as I know, but anything can happen.
What's really sad is how commoditization stifles innovation. Web services expect certain information to be available, certain functions to be performed and (especially) certain timings to be enforced. Web services requests all run according to the clock. Mustn't keep the rock star waiting! So even though some will claim there's still room for innovation in "the plumbing", that innovation can only support two results - a smaller form factor (especially for hardware but also for the software "footprint") or a lower cost. Innovations offering additional functionality need not apply. The new, fun, exciting stuff has to be left to the rock stars. How very, very sad.