cool, would love to see it running on a handheld!
voxel
Creator of
Recent community posts
Thanks for the well reasoned feedback!
The balance between lookahead, speed and cornering ability is a tricky one, and one I tried to mitigate through telegraphing the nature of upcoming turns through the diagonal stripes across the track, most obvious on the first track. If the player can learn to read and follow those stripes, no memorisation is required. It's certainly not perfect, and the game would no doubt benefit greatly from being more explicit about this element.
I'm using https://github.com/KallistiOS/KallistiOS which is a FOSS reimplementation of the dreamcast SDK, paired with https://github.com/Kazade/GLdc which is a performant opengl implementation (but incomplete and with a few bugs). So I get to use opengl1.2 for console dev which is pretty great!
There's documentation for most of this (though a lot of it is out of date) and KallistiOS even comes with a script to automate 99% of building the cross-compilation toolchain.
Thought I'd start a thread to share my progress as I go. My project is targeting some late 90s hardware, so I don't have access to shaders and my rendering budget is very limited, so I need to keep polygons few and blending minimal.

Beginning with the traditional single triangle to check rendering is working

A wireframe grid to represent the ocean surface

Some (animated) displacement to simulate waves moving

Finally apply a texture in the style of Zelda Wind Waker. It's very ugly in the distance without mipmapping, but this is a problem for later in the jam. The white square is a test object that bobs up and down with the water to test that I'm calculating the height of any given point correctly to allow for things to float on the water later
Thanks for playing @Zackarotto. Sorry the game made you queasy, that's never a good outcome. Interested to hear why you didn't feel the floaters were stealthable - our intention is that they have a predictable behaviour (turn side to side) and that their field of view is visualised with a light cone to assist the player in avoiding it. Do you have any suggestions for how we could we communicate that better?
It's not open source, but the first iteration of the engine was closely based on https://lodev.org/cgtutor/raycasting.html
The NES dpad is four 'buttons' for a total of four inputs, so you could argue mouse movement is four inputs. But also there was the light gun for the NES https://en.wikipedia.org/wiki/NES_Zapper which could be used to point and 'click' any spot on the screen, so its not like using mouse input is completely crazy, just pretend its the light gun
Hey everyone! Excited to see what you all come up with for this jam. For myself, I'm planning on making a small NES rom to run on real hardware. I'm not sure how much time I have so it might be as minimal as a static screen with some beeps, but hopefully I'll be able to get a simple game running before the time limit :)