Saturday, November 19, 2016

Backwards

It is easy to get lost when tackling a difficult problem. It is easy to go off on a dead end, wasting a lot of valuable time. When we are presented with a lot of choices, it is very difficult to pick the right one.

For whatever reason, most people are far more comfortable with approaching problems ‘forwards’ from the top down. However, we could see the path to a solution as winding through a rather dense tree of decisions and effort. At every opportunity there are plenty of choices, options. If you make the right choice, you move to down to the next level of the problem. If you pause or head down an irrelevant path, well…

The most optimal way I’ve found is to visualize everything backwards. That is, instead of starting at the beginning, you go straight to the end. If you understand what parts of a solution ‘must’ exist, then you can work backwards to bring them together. Along the way you can collapse any special cases to simplify them. As you keep going backwards from the end, you are naturally pruning off all of those dangerous decisions, all of that useless work. You are avoiding massive amounts of wasted time.

When the outcome is predetermined, the path between there and here is straightforward. If you can walk just that path, and only that path, then you can create a solution in the fastest time possible. All of the work to be accomplished has a very specific, and targeted use, none of it is wasted.

Software gets a little more difficult because it isn’t just one version, it grows over a long period of time often changing directions a bit. As you build a system, there are usually multiple destinations at once and with usage new ones become apparent. Still, if you go backwards from each, and you understand the overall territory you are moving through, then you easily become aware of multiple overlapping paths. You can take advantage of these to further optimize not just the next solution, but a whole series of them. The better you become at identifying viable opportunities to leverage existing work, the more ambitious you can be to tackle bigger and more interesting problems. Time is always the one resource that a Software Developer has far too little of, so saving huge chunks of it is way better than trading off little chunks now for massive problems in the future. The direction of a big project is never arbitrary, and if it seems like it is, then it is only because people haven’t spend enough time to really think it through. If you don’t get that time, because you are wasting it on unnecessary decisions and dead ends, then you’re squandering your most precious resource.


Most development projects are not nearly as difficult as they seem. Most of the endless series of decisions we make are not really necessary. Much of the work done has no real practical value. We get into time crunches most often because we are not spending our time doing valuable things. While often it's hard to see the right path going forward, it is much easier to see it if you are backing up from the solution. All you need to do is flip your perspective and the rest comes naturally.