Search found 4 matches

by ashkatchap
Sun Apr 14, 2019 8:32 pm
Forum: Questions and Help
Topic: MPRToolbox precision
Replies: 6
Views: 338

Re: MPRToolbox precision

After a lot of work implementing SAT and learning from it (and learning that even with edge culling using GAUSS it is still slow for convex-convex, but fast for spheres and cylinders, at least faster than MPR) I remembered something from your useful post, so after coming back here and reading it aga...
by ashkatchap
Thu Mar 21, 2019 12:15 am
Forum: Questions and Help
Topic: MPRToolbox precision
Replies: 6
Views: 338

Re: MPRToolbox precision

Really thank you a lot, for such a detailed explanation and your solutions present in v2. I want to try v2 in my computer, but both of them (one with an integrated gpu that can only run dx10, and the other with a more modern Radeon R9 380, both using Windows 7) are unable to run the demo because of ...
by ashkatchap
Sun Mar 17, 2019 5:04 am
Forum: Questions and Help
Topic: MPRToolbox precision
Replies: 6
Views: 338

MPRToolbox precision

Hi I'm learning physics programming using this library and others, and I encountered problems with MPR. Afaik MPR should be correct enough for any shape pair as long as it is not too big (from the point of view of a float and operating with it). I changed slightly some files so that all collisions a...
by ashkatchap
Sun Feb 03, 2019 4:59 am
Forum: General
Topic: Deterministic fixed-point port, open source
Replies: 28
Views: 4048

Re: Deterministic fixed-point port, open source

I think I just fixed it (I will make a pull request asap) (edit: PR here https://github.com/sam-vdp/bepuphysics1int/pull/1) The problem was here: Fix64 denom = F64.C1 / (va + vb + vc); simplex.W = vc * denom; simplex.V = vb * denom; Changing it by: Fix64 denom = (va + vb + vc); simplex.W = vc / deno...