Michael Dawson-Haggerty

Results 107 comments of Michael Dawson-Haggerty

Yeah that would work, the point size is hardcoded in the windowed here: https://github.com/mikedh/trimesh/blob/64a61ea8ee1695db81ebf9a36b1ed57c3e79e4c1/trimesh/viewer/windowed.py#L409-L410 That could probably be made an argument passed to the viewer pretty easily (happy to take...

No worries I run into this too on high-DPI machines. I made it settable in https://github.com/mikedh/trimesh/pull/997/commits/6913a7ef0031695c5da3287cb3a8a9cc842cde27 and also set the default to a fraction of resolution. Thanks for the report!

That would be awesome! I think this branch has some nice stuff (parent class for intersector objects, kwarg rename with deprecation warning) but I didn't have time to get it...

Sorry for the delay, thanks @aluo-x and @sampotter for the help! Yes that sounds terrific! > We would likely need to distribute embree ourselves, there was a UV bug in...

Awesome! No hurry. Oh nice yeah those Optix bindings look like they might be a little less heavyweight than some other GPU stuff.

I totally "lost the dot" on this, need any help getting this over the line @aluo-x? Happy to help with intermediate results haha.

No hurry just curious! There's a few open ray issues that seem like they might be resolved by embree3, specifically the precision and "missed hit" issues.

> This pull request seems like an interesting and important project. I have been recently working on speeding up my raycaster using the [NVIDIA warp libraray](https://github.com/NVIDIA/warp), with outstanding results. Even...

TODO: - [ ] fix embree3 binding - [ ] re-factor parent to be more scene-like, `RayScene` and then `RayMesh` just wraps the single-mesh case. This was originally written a...

Have you looked at [mesh.outline](https://trimsh.org/trimesh.base.html#trimesh.base.Trimesh.outline)?