Instead, the thing I was working on took longer than expected, and I was finding and fixing bugs in other people's code, and the next thing I knew it was 6:30am. So I figured at that point, with dfan due at 7-ish, I'd stay up and, even if I was too tired to code, I could give dfan feedback and explain my code and suchlike. Instead I actually ended up doing a fair amount of coding right up to the deadline. I haven't done 24 hours straight coding in a long time. The occasional all-nighter, sure, but without any significant breaks? No. (Remind me someday to tell you about the time Doug Church and I spent 40 continuous hours without sleep trying to fix a bug.)
At around ten or eleven pm last night, I realized I was at the 12 hour mark, and I was staring at code and not actually getting anything done, and I said, I should take a break, maybe for an hour or two. But then I came up with a really neat idea, and my teammates were enthusiastic about it, and so my not-quite-up-to-par performance was forgotten as I struggled through. Something around 2am I got a second wind and my productivity grew. From 6-8 am I didn't do much but stare at the code and review.
It may be a month before we hear results. We're not expecting to do particularly well, because we chose to use a programming language (Ocaml) that none of us were very familiar with. But it was fun.
Now I have ten million other things needing doing. But for now I'm trying to stay up so I don't horribly night shift. I have 5 DVDs piled up; two down and three to go.