Stories about Software


Decision Points in Programming

I have a sort of personality quirk that causes me to constantly play what I describe to others as the “what-if game.” This is where I have some kind of oddball thought about altering something that we take for granted and imagining how it plays out. Lest you think that I’m engaging in fatuous self-aggrandizing, I’m not talking about some kind of fleeting stoner thought like “what if I had like eight million Doritos and also X-ray vision?!?” I mean that I actually really start to think strange things through in detail.

For example, not too long ago I was in an elevator and thought to myself, “would I ride this elevator if I knew that there was a 1 in 10,000 chance that the elevator would plummet to the bottom of the elevator shaft?” I thought that I would. I was going up a ways and the odds were in my favor. I then thought to myself that this was a rational choice but viscerally insane — why take the chance?

This led to the thought “what if elevators around the world just suddenly had those odds of that outcome for some reason, and it was intrinsic to the nature of elevators?” Meaning, nothing we could do would possibly fix it. Elevators are now synonymous with 1 in 10,000 plummets. How does the world react?

It’s a wild thing to think about, but the predictive possibilities and analysis are endless. First of all, we’ve got all of these tall buildings, so it’s not as though we’d just leave everything in them and become brownstone dwellers. Some brave souls would go up to get things from these buildings and keep playing the odds. The property value of high-rises would immediately plummet, and you’d probably invert the real-estate structure nearly overnight with suburban/country home prices skyrocketing and swanky downtown high-rises becoming where extremely poor people and drug addicts lived (who else would routinely brave the odds?) I think the buildings would still stand because of the sheer amount of elevation required to knock them down and the fact that we actually develop quite a tolerance for risky things (like driving to work every day).

There’d also be odd anthropological effects. I’d imagine that a whole generation of teenage thrill seekers and death defiers would start doing elevator joy rides to prove their mettle. People would develop all kinds of cargo cult ways to stand or sit in the elevators with a mind toward simply surviving the plummets. In fact, perhaps humankind would just become really good at making the plummets survivable. Politically, I’d imagine that a huge wedge issue debate would emerge about freedom to ride elevators versus the sanctity of life or something. I could go on forever about this, but I’ll have mercy and stop now.

The point is that I take these mental trips several times per day, considering a whole variety of topics. Most of the thoughts that emerge are bizarre and beneficial only as exercises in creativity, like the elevator example, but some are genuine ideas for reboots in thinking about our craft. I find the exercise of indulging these mental divergences and quasi-daydreams to be a good way to get the subconscious brain working on perhaps more immediate problems.

So if you’re up for it, I invite you to have a go at this sort of thinking, but perhaps in a more structured sense. At times in the history of programming, decisions were made or ideas proposed that wound up having a profound effect on the industry. Imagine a world where these had gone differently:

  1. Tony Hoare introduced what we later called “the billion dollar mistake” — he implemented the concept of 0 as a null reference. But what if there were no null?
  2. A lot of what we do to this day as programmers has its roots in decisions made for the typewriter: for example, the QWERTY keyboard and using CR/LF for end of line. What if these conventions had been different when the computer started to take off?
  3. Edsger Dijkstra famously swung the tide against the use of GOTO as a programming language construct with a seminal paper. What if it had popularly stuck around to this day and GOTO statements were still something we thought about a lot?
  4. Of the three programming paradigms (structured, object-oriented, and functional), functional is the oldest, but it lied dormant for 40 years or so before gaining serious popularity today. What would the world look like if it had been the most popular from the get-go and stayed that way?
  5. C++ really took OOP mainstream, but it did it in a language that was effectively a superset of C, a non OOP language. This allowed for the continuation of a very procedural style of programming in an OOP language. What if that cut had just been made cleanly?
  6. What if the most popular object oriented languages didn’t have the concept of “static” and everything had to belong to an instance?
  7. What if Javascript had been carefully planned in an enterprise-y way, instead of thrown together in 10 days?
  8. If disk space had been as cheap as it was and the need for stored information rather than calculation had been higher, would the RDBMs as we know it ever have become popular?

Thinking through these things might just be a random exercise in imagination. But, who knows, it may give you an oblique solution to a problem you’ve been mulling over or a different philosophical approach to some aspect of programming. Things that we do, even highly conventional or traditional ones, are always fair game for reevaluation.