That very week I switched my major from physics to computer science. The code I wrote then was truly, truly awful[1], but the feeling it brought the first time a line of cubes was correctly drawn was incredible. I, and only I had made it do that.
I've not felt that way again. Programming contests bring only the feeling of excelling in competition. Algorithm textbooks bring only the short-lived satisfaction of having a magic trick explained to you. Now it seems any time I research a new project I'm drowned in best-practice recommendations and intense fear of writing my own extensive library when a tested, vetted, and fully functional copy exists elsewhere.
What to do? I would do anything to recapture that feeling.
[0] http://i.imgur.com/wih35.png (find the tetris pieces!)
[1]* https://github.com/ahelwer/UofC/blob/master/cpsc217/as3/grap...
* If you want to run the script, you'll need the modified turtle graphics module found in the same github directory.