One thing I liked about the article is the breakdown of programmers as either engineers or scientists. They say that physics people make some of the best progra

One thing I liked about the article is the breakdown of programmers as either engineers or scientists. They say that physics people make some of the best programmers and from what I've seen it is because they adopt a 'scientific attitude' towards the systems. I don't mean scientific in the sense of traditional CS but rather the ability to quickly devise experiments that offer the most information for the least amount of work. When you come at a modern IT system you have have to accept that you won't have the time or resources to understand most of it at the source code level. It is better to be able to quickly experiment and synthesize the results to arrive at a working model of what is really going on. I think people with science (or even philosophy ) backgrounds can be better at this than people who did an actual CS degree.