What is a Service?

Recently, when I was asked to define the term Service I couldn't come up with a decent definition. In my head there were was this confusion of different definitions, beginning with the W3C defining a Service from a mere technical perspective and some other definitions like we have in S-Cube. On top, after listening to a presentation by Paolo Traverso who talked about physical Services which execution take longer than the execution of software (virtual) Services and human provided Services (which hide behind a machine readable interface), a clear and simple (one sentence like TCP is a protocol stack which...) seemed to me farther away than ever. One finds him/herself reminded on the joke about three business men having four different opinions about the same thing if one happens to be a Harvard guy.

The one million dollar question is obviously: well, why is this the case? And if we know the cause of this confusion, can we learn from it and come up with a accepted definition of Service? I speculate that we do not really have to. Why? Because all over different disciplines there is no broad consensus how central terms are defined. Take for instance biology and genes. Believe it or not, a exact definition is still a matter of dispute among bio tech scientists.

This seems to be a pattern which can be observed throughout scientific history: classification is usually the most tricky part and often changes because of discoveries: before knowing about genetics, animals where classified (or defined) according their external attributes, but the discovery of genes paved the way to different classifications.

So, what do we make of this mess of not really being able to define what a Service really is? I'd say we make the best of it: we discuss it. We argue. We learn from others. We write papers. That's what we do. So the next time, if someone asks me what a Service is, I will be prepared. At least, I'll be able to ask a counter question: what do you think it is?

 

- Martin