Evangelizing Mainframe
Print Email

Non-Linear Thinking

If I were to ask you or any other group of people what your brain is for, you would probably reply that you have a brain so that you can think. It seems pretty obvious really. Without a brain, no thinking whatsoever would take place.

But is that how the brain evolved? Did some creatures find one day that they had a structure that allowed them to think and it gave them an evolutionary advantage over the other? Yes and no. That structure giving them success in the survival of the fittest had to come from somewhere—and that somewhere seems to be as a way to control movement. It’s like animals evolved a coxswain (using a rowing metaphor here) to ensure everyone (i.e., each part of the body that could effectively move) rowed in time to maximize their speed through the water and control the direction they went in. Over time, the cox evolved to become the captain of the ship. You might be interested to know that some adult sea squirts stop moving (become sessile) and digest their own brains because they don’t need them anymore.

That, in part, explains why moving—a brisk walk or a trip to the gym—is good for our brains as well as our bodies. But, clearly, our brains do more than control our movements; they allow us to think. And yet, it’s very easy to not think very hard, to rely on habits. It’s a bit of a trade-off in evolutionary terms. The more we think, the more food and oxygen we use. (About a fifth of all that’s used in the body gets used by the brain.) So if we can get by without thinking too deeply about any decision we make—like should I eat that last donut or should I stick to my calorie-controlled diet—the fewer calories we burn. This energy saving was a good thing for food impoverished pre-humans, but makes no difference in 2016, when we can simply pop to the supermarket.

Let’s look at some examples that don’t need too much thinking about. If I can drive 30 miles on one gallon of fuel, how many gallons will I use up to drive 60 miles? The answer (and there’s no prizes) is two gallons. And if I need to drive 120 miles, I need four gallons of petrol. This is an example of a straight line graph – as the distance axis increases, so does the amount of petrol used. And we tend to use this kind of straight-line thinking to solve lots of problems because it’s easy and because it is so often correct.

Here’s another example. If I drink one pint of beer (or glass of wine) I get one unit of pleasure—let’s not worry too much what a unit of pleasure is, you get the idea I’m sure. So, as I’m finishing off that first beer and someone suggests having another one, using my straight-line thinking, I would assume that two beers would give me two units of pleasure. Therefore, how much more pleasure would 12 or 24 beers give me? The trouble is, that once I start having a beer, my memory of those occasions when I’ve suffered from a terrible hangover the next day or, worse, have spent the evening being ill into the toilet, disappear, to be replaced by memories of me dancing incredibly well, and being the life and soul of the party. (Remember, I said they were my memories, not necessarily shared by anyone else who was there at the time!)

What about applying our straight-line thinking to the cost of running an IMS or CICS system on a z/OS mainframe? Does it work? Let’s suppose that we have the initial cost of the mainframe (we’re renting an older model, but we could have a top-of-the-range IBM z13). And there’s the cost of the software (z/OS, DB2 perhaps, IMS or CICS), there’s staff costs, and buildings and utilities (heating, lighting, etc.) costs. That’s basically our overheads. If we just turn on the machine, but don’t run any transactions, there’s a cost. If we calculate how much per transaction this is costing us, then it’s infinite. So let’s start running some software. The cost per transaction starts to decrease, which is good. But if we keep loading and running transactions, we’re going to find that IBM starts to charge us more. So our costs start to go up again. What we have is a bell-shaped curve. It doesn’t have to always go up in the middle; it could start high at both ends and reach a low point in the middle. And it’s not always symmetrical, it can be skewed to the right or the left.

What I’m suggesting is that many things in the world (like the best price to sell coffee: too low and you don’t make any profit, too high and you don’t sell any coffee) follow a bell-shaped curve rather than a straight line. It’s very easy to think of things in terms of straight lines, even when that’s not how things work. You may hear management say that losing one of the remaining mainframe staff will save us x dollars, therefore getting rid of all of them will save us a multiple of x dollars. That’s clearly an example of straight-line thinking and so obviously wrong.

There’s an optimal point on the curve to determine the best price for coffee in a neighborhood. Similarly, there are an optimal number of people to be working on the mainframe at your organization and an optimal amount of experience to have in your mainframe team. Encourage your management to use their brains and get away from simplistic straight-line thinking to get the most from your mainframe and give your organization a competitive advantage.

Trevor Eddolls is CEO at iTech-Ed Ltd, an IT consultancy. A popular speaker and blogger, he currently chairs the Virtual IMS and Virtual CICS user groups. He’s editorial director for the Arcati Mainframe Yearbook, and has been an IBM Champion every year since 2009.

Posted: 6/14/2016 12:30:14 AM by Trevor Eddolls

Print Email

Join Now!