Another Project I Probably Will Abandon



  • So lately I was reading a thread here, listening to people talk about generating dungeons, and how they shouldn't be perfect cubes all the time.

    So I started to wonder if I could do something like that, and generate (at first) some kind of overworld map.

    I decided I'd do it with extreme TDD for practice, and go ahead and use Fakes on it. I've only let myself create one factory class so far, but I intend to go as deep into Immutability Land and SOLID as possible. It will definitely be overly complicated.

    I'll probably post code examples or something eventually. I haven't worked on it all week, and I'd barely started anyway.

    I intend to at the very least generate maps that have oceans and lands split into countries of varying size, with buildings or something in them.

    Since I'm writing this primarily as a class library, I'm slightly at a loss as to how to store countries etc, because if I ever want to output to a console, I'd have to do something horrible. I could do some kind of vectors easily enough anyway, though...



  • Clearly you need to use a spatial database!

    Filed under: Contents may or may not be sarcasm





  • I remember that thread. It was the one where I said there was stuff I didn't want in the OP, and everyone spammed it at me every post, like with Blakey and CLI.

    Yes, probably. I even named this thread after the fact that it will likely be abandoned.

    Luckily, I can actually write code that does things in C# without being spammed for a hundred threads on how I need to use EMACS first.

    This is also in general rather than help because I don't really mind if people just comment on the idea and take the thread sideways. But help coming up with ideas would be cool too.



  • @Magus said:

    So lately I was reading a thread here, listening to people talk about generating dungeons, and how they shouldn't be perfect cubes all the time.

    You mean the MUD design thread?

    I was suggesting ditching grid-based, because it's dumb and unrealistic. I wasn't suggesting procedurally generating the world, that's a tough nut to crack.



  • Exactly, all of that.

    I don't intend to be very precise on how I do this. For instance, have no intention of making buildings structurally sound when I get there, though they'll probably be 3d, or anything except their size, contents, and position pretty much.

    Should be interesting. I'll probably just go with vector stuff, since I understand that passably well.



  • Well the interesting thing in a MUD context is getting that 3D world and converting it to textual descriptions of a particular place.

    The good news is that since you're generating textual descriptions, your 3D positioning doesn't need to be exact and for the most part you don't need collision detection (although you do need to keep track of what surface the entity is standing on, assuming it's standing on a surface.)



  • What I'm making isn't necessarily for a MUD. It could be used in one for sure if it works out, but ideally it could also be used for other things too, since it's just a world generation library.

    I don't know exactly what I'll do with it, since it's mostly just me practicing things.



  • If you have no end goal in sight, how do you know when you're done?



  • This is a project for fun, to practice TDD. I'm not terribly serious about it, and don't mind if it's never 'done', but at the same time if it goes somewhere I know I'll have uses for it.


  • BINNED

    @Bort said:

    Weren't you the guy that asked for help on a lisp project but then argued with us for the whole thread and then the project went nowhere?

    Actually, he mostly argued with me for the whole thread.


Log in to reply