I don't believe I can prioritize these between more or less important things, but this is the set of things that I usually credit: * Programming competitions.

I don't believe I can prioritize these between more or less important things, but this is the set of things that I usually credit: * Programming competitions. The main utility I got out of this wasn't learning obscure data structures or algorithms (which tend to often not be very relevant), but rather the experience in coming up with corner cases that might break my own code, and how to write and debug code without sitting in front of the computer. I guess in short it's the ability to separate the algorithm from the implementation and track down failures to one or the other. * Going to school and getting a CS degree. Learning in general is useful, but I believe that the structure of a full CS degree is generally better than picking things up ad-hoc, especially in guiding you towards areas you wouldn't otherwise think to learn about (e.g., I wouldn't have picked up information visualization on my own). * Seek out and learn as much as you can in general. You never know when randomly learning some topic in some distaff field might come in useful; one time, I happened to discover (while looking up information on ptrace in a debugger-like side-project I was working on at the time) that Linux supported hardware watchpoints via perf... and a few days later, we had to resort to that to track down a bug that wouldn't reproduce in gdb. * Peruse other projects for how they do stuff. Part of the above, but you'll pick up on "huh, that's an interesting thing to do" as you learn about how they work. If you're curious about how something works, just open up the source code to see how it works. At the very least, doing so frequently will give you a good sense of how to quickly find stuff in large, unfamiliar projects, which is apparently not as common a skill as I take it to be.