Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Knight of Fools]

I don't know much about programming, and I don't know how you've got it set up currently, but could you have an assigned number for each variation and just have a list of compatible variations for each creature? That way, each creature is only checking against a small list of numbers instead of a whole list of compatible biomes.

So, Creature A is can live in Types 1, 2, 4 and 8 biomes, where each number represents an entire list of characteristics, such as elevation, plant life, temperature, humidity, nearby water, soil, food availability, and whatever. If it's checking against a biome type 3, then it simply runs checks to see if three is in that list. If it's not there, the creature doesn't move in. If it is, it spreads itself out more.

I'm not too sure how well this would work, because you'd end up with a lot of possible combinations, but you wouldn't have to have the program scanning through the other conditions to see if enough conditions match up for that creature to move in. You'd just have one match to make, comparing the biome against the creature, and it's finished.

The pain would be setting it up - If it even works better than what you already have.

[Lord Dullard]

I got the algorithm working much better, although I still want to tweak it a bit because I suspect it will be painfully slow on older computers. It's still much less time-consuming than it was by an order of magnitude or three.

Also, I managed to begin work on the code that handles starting up a new game after a world has been generated. I'll post more updates as my schedule allows. I have double shifts today and tomorrow, so we'll see how that goes (I really need to find a job with a slightly less crazy schedule).

I'm actually really excited to begin work on the parts of the game that will allow real interaction with the world. I have some ideas about implementing the various types of landscapes that I'm itching to try, too. Although this is going to take a while, barring any unforeseen changes in my schedule.

[Sentientdeth]

New track, check out my blog.

[Epithemius]

Sorry, I realise I'm breaking the whole "necro" thing, but has this gone dead? I hope not, it seems like a really great project.

[Servant Corps]

Development has been slow anyway, so I don't think this project has been abandoned yet.

[Lord Dullard]

Nope - not dead. But a bit stalled, yes. It isn't the first time development has stalled; I can't afford to make this a full-time activity right now. I've been working six-day weeks with at least four or five of those being doubles, so... doing anything mentally taxing on my one day off has not really been a priority, unfortunately.

On the bright side, I'm now going down to a schedule of only four days a week, starting this week. Theoretically this means I'll no longer be devoting all of my time off to cleaning, chore-running, and doing other necessary things.

[Sentientdeth]

Woot!

[Max White]

Hey!
It's funny that you drew my attention! You see I am also making a roguelike, with libtcod, although mine is called 'Cultist', 
Anyway, your game looks spectacular. No seriously, few roguelikes qualify as eye candy, but this is great. Looking forward to seeing where this goes.

Anyway, I'm having sort of an issue that might have already been solved, or maybe not, but your window is too big for my screen. I know libtcod makes fullscreen very easy, so if there isn't yet a solution for this then that is something to look at.

Either way, good luck, and happy coding!

[Lord Dullard]

I will add a fullscreen setting to the game at some point. The code itself already accommodates (to some degree, at least) user-specified resolutions set via an init file, but I haven't bothered to bugger around with any of the init stuff as long as I'm just doing early dev work. Since I hope to have some gameplay elements in place for people to test with the next release of the alpha version, I'll probably include an init file with it that'll allow users to either specify their own resolution or switch to fullscreen if they prefer.

Oh, and keep me updated on your RL.

[Lord Dullard]

Well, I've hit a bit of a coder's block.'

I'm starting on town generation for when you embark into the early parts of the game (for the purposes of simplicity, at least for now, the player will be selecting a city or village to begin the game in, probably in a somewhat Darklands-esque  manner). But... I really don't know where to start with procedural generation. That is to say: should I start with the terrain, and just work from there? The problem with that is that I'd have to code in all the settings for multiple terrain types, which seems a little over-ambitious for the moment. Also, I'm not exactly certain how I'm going to have the world generate map edges that match up to each other (i.e., if you wander across the border of one map square and into the next, how will the game make the terrain scroll smoothly while procedurally generating each square?).

So... yeah. Just a bit stuck, at the moment...

[Sentientdeth]

Feeling the same way right now.  Think I've got the winter-dontwannadonothin-blues.  Probably has something to do with me taking most of my inspiration from nature, and it's too fucking cold to go outside.

[EagleV]

Assuming you use perlin noise, say you've got a tile with height 5 next to one with height 4. Then on tile 5, the border with 4 will get a low value automatically for the least detailed noise function, and random values for the others as usual.

Example: say your biggest noise function has an amplitude of 32, your tile height 5 means it starts at 500m, 4 means its start at 400m.

The tile 5 borders would then have a height of 500 - 32 (lowest possible) +- 31 (due to the other, still random perlin functions) m.
The tile 4 borders would then have a height of 400 + 32 (highest possible) +- 31 (depending on neighbouring tiles) m.
You can then make sure the tile 5 and tile 4 borders don't differ too much, by adjusting the lesser perlin functions in the borders of tile 4 (the underlined factor).

As for fitting buildings in terrain, you could make a few functions, 'build_on_flattened_ground', 'build_on_poles', 'build_in_hillside', then choose a few random places for buildings with higher chances on flatter ground or near water, and then, flatter ground has a higher chance of spawning a normal building where the ground is leveled first, hills have a higher chance that the building will be partially dug in or with a side that rests on poles.

No idea if you can do anything with any of that, but good luck!

[Max White]

Well the first step is to try and define what a town is.
What does a town have in it? What is absolutely required? What is optional? Think about every possible thing that a town has and list them down. Buildings and subtypes of buildings, like houses and pubs, roads, people, wells, walls, trees, all of it.

Now that you have what you need, think carefully about how they all join together. Plot out a mind map if you need, but remember that how you structure this will be one of the biggest aspects of how your towns will look. Start with the concept of a town in the middle, and go from there.

For example, if I decide a town has buildings, and roads go between the buildings, then that would imply genning a bunch of randomly placed buildings and fitting roads between them. This would cause a great little maze of buildings full of side allies and hard to navigate.
If, instead, I decide that a town has roads and buildings are placed on the sides of roads, then I would gen my roads first, giving me longer, straighter roads that are easier to navigate, and neat little rows of buildings lining the sides.

Just think about what you want to achieve, and break it down into smaller steps.

[Quarynn]

*drools* World generation...

[Lord Dullard]

Well... finally making some progress. I just needed to do something to get started.

Thanks for the input, guys.