Cryopyre wrote:I just tried that but realized I might be missing dirextx 9.
Edit: It's working, it's definitely got potential. The problem is my computer, though pretty good, it crashed it quickly after starting, and when it wasn't crashed, it was moving incredibly slowly.
Try holding '1' as soon as it starts. You won't be able to see as far, but it should run a lot more smoothly (recursive depth defaults to 3). '0' will work as well, though then you will only see the room you are in.
In terms of performance, here is why a depth of three is so costly...
Code: Select all
int num_rooms(int depth)
int result = 1;
if(depth > 0)
result += num_rooms(depth - 1);
result += pow(6, depth);
This evaluates to 1 + 6 + 62
= 259 rooms rendered, versus only 7 at level one. There is also an overhead of 62
render to texture operations, as every other recursive step it becomes possible for multiple different rooms to occupy the same 'unfolded' 3D space, and which room you see must depend on which walls you are looking through. At the maximum depth I support (a depth of five), there can be as many as 9,331 rooms, and 1332 render to texture operations!
There are some optimizations in place: Detailed wall geometry reduces to flat textured planes at depths greater than one room away, walls which face away from the camera are not recursed, and view frustum culling is applied to further reduce recursion when walls can not be seen. Still, it just isn't enough to pull off a steady frame rate when you can see really far in every direction you look.
Cryopyre wrote:SO what are you working on with it now?
We are preparing to break into two 15-person teams at school to build Half Life 2 mods (these will be twenty week projects). Each student was given the opportunity to give a 5 minute game idea pitch last week. I pitched the tesseract. Thanks to the demo it made the top ten, and gets to be further fleshed out and re-pitched for a longer time this Friday. Unfortunately, I don't get to be the one to pitch it.
For some reason, the prof thinks our ideas will be vastly improved if the original authors are shut out of the loop and a group of three other random students get to apply their own spin. In addition, the prof combined my idea with someone else's, which was basically 'you can flip gravity any direction whenever you feel like it'. This would destroy any hope for building interesting puzzles within the tesseract. Since it is such a challenging environment to wrap one's head around, I fully expect the group that is pitching the concept to focus on the gravity mechanic and completely nerf the 4D idea. I really hope that this doesn't happen, but if it does then the tesseract project may have to be put on the back burner for a long while.
In the mean time, I get to work on a pitch for a game where you poke an imp to try to help it escape from a dungeon. Le sigh.