Deku's Development Blog

Sliding

noreply@blogger.com (dekutree64) — Sun, 20 Jul 2008 23:36:00 +0000

Ok, I think I'm done with collision for now. It could use a bit more tweaking to prevent you from walking up nearly vertical walls and such, but I don't think I'll need any non-vertical walls anyway.

I'm not really sure what to do next though. I suppose it's time for a list.

- Character structures and a list for updating them.
- Build an enemy model, with idle, walk and shoot animations.
- Shooting controls and hit detection (so much for being done with collision).
- Proper velocity based physics and jumping.
- Health and ammo pickups.
- Some way of defining enemy and item placement in levels.

That ought to be enough to stay busy for a while.

Walls

noreply@blogger.com (dekutree64) — Sun, 20 Jul 2008 07:58:00 +0000

I now have walls blocking movement. I ended up spending a good chunk of the day scrambling around with more sphere vs. triangle code because it wasn't working quite right before. Now I'm into the higher level collision response code and management code. Nothing too difficult, but fun to work on so I'm making quick progress.

Small progress

noreply@blogger.com (dekutree64) — Mon, 14 Jul 2008 08:50:00 +0000

Well, so much for setting weekly goals. It turns out collision detection is still hard. I think I've got the basic sphere vs. mesh check working though, so it shouldn't be long before the ground becomes solid.

First post

noreply@blogger.com (dekutree64) — Mon, 07 Jul 2008 09:30:00 +0000

So about a month ago I was bored and decided it was time to start up a new weekend project. I started tossing around game ideas, and ended up settling on a first-person shooter in the vein of Turok: Dinosaur Hunter. A sort of jungle/ruins theme, and relatively dense fog for artistic effect (though maybe not as opaque as Turok's, since I don't need to worry much about speed), and probably a fixed low resolution (320x240 or maybe a bit bigger). I think I can pull off the atmosphere I'm going for much better than if I'm worrying about how it will look at 1680x1050. Beyond that I haven't made many solid design decisions yet.

I've been coding away since then, and loving every minute of it. I decided to do the whole thing in C, as I prefer the style of code it tends toward to that of C++. I'm using SDL, with OpenGL for rendering, and other than that starting from absolutely nothing.

My basic sequence of attack so far has been:
1. Write lots of math functions.
2. Get a grid rendering, and a basic camera system set up.
3. Buy an XBox 360 controller, because it will make working on the game more fun.
4. Write a fancy input system so I can map controls to different devices.
5. Get 3D models loading from my favorite 3D modeller, trueSpace.
6. Load animations too.

All of that went fairly smoothly, except for a bit of suffering at step 5 due to confusion between how my matrix math worked, vs. what OpenGL wanted, vs. what the model data was giving me.

Today I've been working on collision detection. Good fun, as I've done it plenty of times before. I've decided to go with axis-aligned cylinders and boxes for moving things, and ellipsoid vs. triangle soup for environments. So far I have all my shape intersections and mesh collision data loading done. Next up is running an ellipsoid into that mesh.

Sadly, the 3 day weekend is over, and I must go to sleep. Goal by the end of next weekend: Collide and slide against level geometry and manually created cylinders.