i had previously made a plotting script for my own private leaderboard, here’s the plot for the racket leaderboard if anyone’s interested :slightly_smiling_face: (https://git.lain.faith/haskal/aoc2020/src/branch/aoc2020/scripts/plot)

It seems to me that those on here who are top of the rank use the for constructs more than I do. So I wonder whether that’s a point of improvement for me.

Also, I think I write too much code before writing tests; at which point it’s obviously harder to pin down the problem; and that slows me down.

Does anyone else have “lessons from previous AoCs”?

> for constructs more than I do. So I wonder whether that’s a point of improvement for me. Use whatever works best for you, but I think for is very handy, so I prefer it.

> Also, I think I write too much code before writing tests; at which point it’s obviously harder to pin down the problem; and that slows me down. My opinion is that competitive programming does not encourage good software engineering practices. If you aim to be at the top of the global leaderboard, you need to get it right on the first attempt and don’t waste time on testing. But if you don’t try to participate competitively, testing (early) helps.

Disclaimer: I haven’t participated this year’s AoC (have been busy…). But I was at the top of the Racket leaderboard last year.

for constructs are super handy

but also I like to use old school let loop sometimes as I can exit early

<https://github.com/lojic/LearningRacket/tree/master/advent-of-code–2020|Day 6> brought to you by the threading macro :)

Day 6 was surprisingly easy. That said I could probably optimize a lot, I’ve got a few nested maps

the difficulty for day 7 went a bit up a cliff u___u