Slow start, some procedural generation
reidrac on 2011/09/11 17:52
of usebox.net
— edited on 2011/09/12 08:16
First. I've updated my entry, I have a title and a plot based in the theme (Mutation! yay!): Operation: Cell's Genome. Basically I'm going to implement the game idea I got past weekend having coffee with my partner (credit where is due!).
Second. I've been implementing an algorithm to generate the levels. Actually, it's just a generator for "a set of rooms joined by corridors", but it's definitely a good start.
References:
- Basic BSP Dungeon Generation: I've found this article very interesting, although my implementation it's not 100% compliant with the described algorithm. I'm not using a tree (oh, not strictly at least), but I'm happy with the results.
- Procedural Generation: wow! PyWeek references itself! Worth reading too.
Yes, it's just an ASCII output. No pygame code so far. This "meta" stuff was a good start, it feels good to see some progression. Btw, I'm not doing formal testing besides iterating 10000 times over the code and checking it doesn't crash. Close enough :)
I can tweak different parameters, such as size of the map, recursion level, etc; and the SEED for the generation. So it doesn't need to be random EVERY time. I'm thinking about it.
If you're interested, there's a lot information about procedural generation in rogue-like games (check the wiki in the previous link about BSP).
I'm going to work on some basic tiles fitting my game plot, and start with pygame stuff!
PS: the rich text entry editor for this journal is really frustrating. Can I haz markdown? :)
(log in to comment)
Hugoagogo on
2011/09/11 21:14:
Looks very good, could be some frustrating long chains of trees though, it has a very nice overall fill of the map though