How to think about programming

The question I asked wrong#

The first thing I asked when I started was: "which language should I learn first?"

I spent weeks reading language comparisons. Looking back, those weeks were close to worthless.

The better question is: how do I break a big task into steps small enough for a machine to follow?

A language is only the notation for writing those steps down. Switch languages and the notation changes; the hard part stays exactly where it was.

The three things that mattered most#

1. Breaking problems down. Turning a vague request into concrete steps. This is the skill I spent the longest practising and the one I use the most.

2. Reading other people's code. I used to think programming was mostly writing. In practice I read far more than I write — my colleagues' code, library code, and my own code from three months ago.

3. Debugging with a method. Early on I fixed bugs by guessing and changing things at random. Later I learned to narrow the search: work out what is definitely correct, what is definitely wrong, and squeeze the suspect region until it is small.