Crazy Coder's Corner

Want to hear crazy coder's stories? You're at the right place! Warning, occasional MLP!

[crazycoderscorner]

Collision response

So, one of the problems I’ve met when writing a game engine was collision checking and response. I’ve tried it in 3D but it proved to be too difficult to me (yeah, I’m not exactly sharp). So the solution is - SAT. It’s a 2D algorithm but can be scaled to 3D. Basically you calculate a dot product and based on this calculate how much should the player object be moved back. It’s a neat algorithm so if you’re looking for a collison checking algorithm check it out.

[crazycoderscorner]

Welp

Too much work makes me crazy.

Anyway, the game is still a thing and is still probably being made with GM 8.1. While moving to my software rendering engine sounds fine and all, I don’t really want to use it for anything other than experiments.

[crazycoderscorner]

Phew

After a looooong time a little update.

Currently remaking a HTML5 software engine in Java. Looking good so far. Not sure if I’ll move the game to it but hey, at least game concept is evolving. The game will probably (probably) have three short episodes, which will make my work a little easier (making a short episode is faster than making a long game).

[crazycoderscorner]

Moving to GM 8.1

So I’ve decided I need a higher framerate, thus moving to GM 8.1.

Why not Studio? Because I don’t like it’s ideology of “sandbox”. It doesn’t let me use local files without some crazy mumbo-jumbo, so I don’t use it. So, now I have a working skeletal snimation for GM 8.1 at about 120-150 FPS. Also, a lot higher resolution (up to 1920 by 1080 without loosing too many FPS).

[crazycoderscorner]

You know, sometimes you begin to think that what you make doesn’t look good...

I guess that the problem every creator gets sooner or later.

I managed to make some animations but some frames look strange. Need to redo them but I’ll do it later. Maybe.

[crazycoderscorner]

Ugh. Making assets like animation and maps takes so long...

[crazycoderscorner]

Optimization

Soooo... as I said I managed to squeeze at least 20 frames per second from this engine. However while it’s good enough, the more - the better. So I implemented interlaced renderer. That means the engine renders only half of horizontal lines every frame update. That creates “combing” effect but it’s a good enough solution since it’s not that bad. It becomes bad if only quarter of lines is updated though.

[crazycoderscorner]

Dumb HTML5 project

This is the project I’m trying to finish at the moment. I will be posting updates for it in due time. Right now it looks like this:

“Why does it look so low poly and doesn’t have textures” you ask? Well... because this is software renderer made with HTML5 canvas but without Web GL. Hey, I did say I’m crazy, right? It all started as an experiment on what I can get out of Canvas element. Apparently you can make a software renderer work at at least 20 frames per second. And if you set resolution to 320 by 180...

PS1 graphics, yay!

But anyway, the project is about Sunset Shimmer (from MLP, ya), and is going to be a short survival horror game. Hence why I think that simplistic graphics is good enough. It will look worse than Silent Hill 1 (which is the main influence) though. I’d say that in terms of graphics it’s more like Another World in 3D.

[crazycoderscorner]

The beginning of everything

Okay, so... why exactly does this blog exist?

Frankly, I’m not sure myself. Maybe because I want to tell people about my hobby that gets out of control fast.

I am a programmer. And I like making game engines. That’s it, that’s the problem. Not making games, just engines. I do want to create a game but alas only engines came out. I have one or two completed games (which I will be posting here... a little later), but every other project hasn’t been completed yet.

Maybe this blog will help me focus on one project and I will actually finish... something.