/agdg/ - Amateur Game Development General

>>20660

Didn't knew him. This is a great find!

Thank you.

Seems pretty cool. I'd play this, even though top down shooters aren't my cup of tea. I haven't even played Hotline Miami. I haven't really seen horror in this format.

>>20678

- Too much light

- Shouldn't be able to see anything outside the corridor

- Too wide

Make it dark, and cramped. Perfectly repetitive environments can work, but you need to be really careful with them. Normally the environment ought to differ.

Find some way of introducing larger areas for combat, maybe just rooms, or maybe the enemy bursts through the wall and you can exit the corridor into a bigger area like that. Maybe creepyness fucks with space.Maybe the wall on the right turns out to be a part of a huge monster and not a wall, and the area opens up when it moves.

In that fig I can see well and have room to manoeuvre. One feels safe.

>>20845

So how does it work? Is it just putting down random blocks and then reducing it until you get something coherent?

Pretty cool man, it'll be sweet if you make some sort of hive level you have to periodically explore that's randomly generated to contrast a facility like setting.

>>20874

This is a pcg-algorithm called "Cellular Automata". It's pretty easy to do.

And yeah, at first I was thinking of using it just for helping me making map design for caves... Or just a fun random dungeon mode.

But this is a really cool idea!

Please tell me you didn't make that lightning engine, because it looks great. Where did you got it from? Is it easy to use, I want to have proper lightning in my 2D game too.

>>20910

Yeah, I made it. I kinda like making this kind of stuff by my own, so I get better writing my games. The code is pretty bad looking, but it works fine.

http://ncase.me/sight-and-light/

This links explains a lot! If you're in need of doing a lightining engine, that's where I recommend you to start.

Also, there are a lot of free engines online and ready for use at the Game Maker Community.

>>20917

That method is really clever, but it also limits you on the ammount of light sources you can have

because

O(N*I)

where

O is the order of magnitude of cpu time

N is the ammount of light sources

I is the ammount of vertex in the geometry

i preffer a GPU acelerated method

http://www.catalinzima.com/2010/07/my-technique-for-the-shader-based-dynamic-2d-shadows/

it has a O(N) time

is on xna but it works in any computer language where you can use gpu shaders

>>20919

You can even go further and use a bsp tree to limit the ray max distance like in doom

>>20919

>>20920

There are Dynamic and Static lightsources. There will only be the Flashlight as a Dynamic lightsource. Yeah, it's a pretty limiting and slow system depending on what kind of room you have, but it's pretty optimized as it is in my "engine".

I saw this link when I was studying. But I ddn't knew what the fuck shaders were, now that I know some glsl, I'll definitely try in my next project.

Do you know any lightining method that uses Normal Maps?

>>20921

in 2d?, i have never conceived that thought

>>20905

Which cellular automata?

>>20919

>O is the order of magnitude of cpu time

That is not remotely what big-O represents.

big-O is a function which takes a function A (in your case NI) and produces the set {f : ∃c>0:(∃n0:(∀n≥n0(f(n) ≤ c·A(n))))}

>>20983

>Is there more than one Cellular automata?

There is an infinite number, though I'm not sure which infinity.

Thanks for the link.

If you're interested in cellular automata check out rule 110, and conways game of life; they're both turing-complete which is literally awesome.

>Also, "O" represents the complexity of the algorithm.

No, it doesn't, and he isn't.

Saying an algorithm is in O(N·I) is to say that the value of the function f(n,i)=n·i multiplied by a constant when both n and i are at or above constant minimum values is always larger than some metric of your algorithm for the same values of n and i.

In this case the metric was, implicitly, CPU instructions.

>>21021

I've played a lot with conways game of life beforing doing this algorithm.

Rule 110 is really awesome to see! But I don't see any applicability for what I'm doing. Is there any big use for this?

>>21024

>before

>Fix'd

>>21024

It's not applicable to what you're doing, but it's extremely simple and turing-complete which is cool.

>>20660

This plebshit is nowhere near beksinski