Thoughts by Richard Bleil
A couple of years ago, the latest ax to cut me fell. For two years, many of my possessions sat languishing in a storage unit, mostly in Iowa, and more recently in Omaha. Among these possessions are electronic devices that, frankly, I’m not even sure work any longer. As I write this, I’m looking at two desktop computers waiting to make their way upstairs to what will eventually become my office, computers that are probably fifteen years old now.
Yep, I bought them early in the new millennium. One is a Linux box, the other runs Windows XP, and both are laughably old. Like I am, I guess. But there’s a reason for them. See, they both have compilers for my native programming language, Fortran, that people also find laughably old, but Fortran has always been and still is a great language for number crunching. It’s considered a “dead” language because you can’t do graphics or games with it, at least not easily, which is what programming has largely become. The more serious modern languages are geared towards statistics or database manipulation, which is at least serious, but the reason I run Fortran is because I write mathematical models, purely mathematical calculations that have run, in my youth, for several months continuously before coming to a final answer. Yes, months.
The kids who laugh at my antiquity computers and ancient language don’t understand this. The system I’m typing on now is far faster than these beasts, but it will simply not run for months at a time, and while it can do parallel processing, these calculations would still slow it down dramatically from other jobs. And, yes, other computing languages can do pure number crunching, and yet it’s not their primary function meaning it would be just as difficult to get them to do the work I want them to do as using Fortran to make a graphics-based game would be.
Truthfully, I am hoping to get back into programming. I don’t know if or how I will get funding to do it, but there are still projects I hope to work on. For example, I have an idea for crime statistics and prediction. My idea is extreme, and extreme ideas rarely work, but when they do, they usually are game changers. Most crime statistic programs are based on database mining, looking for historical patterns in crime. Mine is not. If it works, it should do everything the current crime programs can do, including creating crime “heat maps”. This will be a great test to see if the program works; by reproducing already accepted outcomes, it’s evidence that it can replace the current methods. However, there is no reason to do this if all it can do is what is already being done, and my approach can. What my program would be able to do is predict crime pattern, and their shifts, if changes are made. For example, putting lighting in a park would, as you might guess, eliminate the crime in that area, but the question becomes how it will affect other areas in the city? My approach will predict how much crime in the park will decrease, and where crime can be expected to increase as those criminals seek new opportunities. It could also play “what if” scenarios. Suppose a city has a limited budget to install crime fighting technology, but would it be better spent on cameras in this location, or lighting in that, or a community garden in the other? With this idea, they could plug each of these ideas into the system and see not only how the crime will change in that area, but how it will affect other areas as well.
The first step is to get the running prototype using a mock city. Once that is completed, then I could put in real data from an actual small city area, perhaps from a town in a state with a “sunshine law”, where crimes are public information. If I can reproduce heat maps, and make predictions on criminal changes, then I might be able to partner with a city somewhere for a live test side-by-side with a traditional crime statistics program, but it all has to start somewhere. We’re talking about a lot of time to get this working, and unless I can find a grant or partner it means I’ll have to do it alone, or at least until I can find financial backing, and it is going to start on laughably antique computers with a dead programming language by an old man alone in his home.
Bleil's Banter 