Just over a year ago, I enjoyed a book by Chris Anderson (editor in chief of Wired magazine) entitled "The Long Tail". In it, the author asserts that we're entering a world of more and more customization or broad variety of available consumer goods and a reduced commerce in commodities. Key reasons for this include the Internet, the availability of e-commerce and just-in-time manufacturing, electronic items (such as digital music) which take negligible space to store and can be digitally copied on demand, and the diverging tastes of the new digerati.

This got me thinking and, deliberate dinosaur that I am, over a year later I reached the conclusion that it's worth blogging about how the mainframe was in many ways the original example of this.

Now, you're probably wondering where the expression "long tail" fits in to all of this. Basically, traditional commerce in consumer commodities, predicated on predictive ordering and storing of what the majority of people would want, can be likened to a bell curve that is truncated where it narrows in order to keep things cost-effective.

But, if the majority can now have their uniqueness affirmed and their unique tastes catered to, suddenly the "truncated tail" of the bell curve becomes where the majority of commerce takes place - profitable even with a single-digit number of items of a given type being sold, due to the sheer number of different, individual items and configurations available.

So, what makes the tail of the "mainframe dinosaur" the original long one then (aside from some spectacular fossil analogies available in museums such as the world-famous Royal Tyrrell Museum in Drumheller, Canada)?

Simply this: that from the very beginning, every mainframe environment has been unique and custom.

Of course, the nice thing is that, unlike the unique and custom hardware and software configurations of every individual PC in most large organizations, each of those organizations usually has only one or a few mainframes, so they're able to build a stable and reliable configuration within their organization - at least until they acquire another company that has a mainframe; then the process of merging to a new stable configuration is needed.

But it's a good thing to remember when thinking such leading-edge thoughts: when you're looking for a good example or insight into where we're going in the digital world, where we've been with the mainframe is often a very good place to start.