Hours of sleep: Around 8.

Lines of code: 2621.

Curious bugs: Innumerable.

Things I did wrong: Almost all of them.

However, I suppose that I did the most important thing correctly, making and submitting “10 seconds for Cactusman” within the 48 hour period.

Looking back upon my development process, I realize that the most blaring mistake I made was made before I began the competition, I failed to write much of an engine, even a rudimentary engine, before the theme was released, and thus spent most of the first and second days writing collision code and various other theme unrelated tasks.

The other mistakes that I made were less damaging. Ignoring the theme until the final ten seconds of development, writing a terrible level-saving format, and finally not correcting bad bugs, but spending hours fixing the most minor of issues, probably set me back a bit, but I honestly don’t know how fatally they’ve crippled “10 seconds for Cactusman” overall.

Anyway, that’s it for my first postmortem, and my first LD.

Hopefully I’ll be able to implement something cooler next time.

P.s. How do you rate games?

After around 8-12 hours of work, I’ve implemented all of the things that I need to for this game, now I can start fixing the odd bugs that have popped up.

The game will probably revolve around having a pool of “seconds” that you may lose in various ways in each level.

Each level will most likely involve some sort of puzzle.

If you run out of seconds and can’t win, you have to start over, in a strange, voluntary permadeath sort of situation.

This is pretty last minute, and this’ll be my first full LD, or LD ever for that matter.

I’ll probably be using python and pygame, or C and SDL.

I’ll most definitely be using sfxr for sound too.

However, as I want to reach the largest audience possible, I’ll most likely use python.

As this might affect my decision, are there any compilers that let you compile to different architecture?

I’m using gcc on Debian if that changes anything.