jmlait

I like the idea of it being built into the CMake, but then it occurred to me that this might be a security issue? It might not be expected that...

As mentioned, we'd ideally have options in Cmake, but we can make that a separate PR if we get there.

The goal, I think, is to handle a situation where openvdb headers exist on /usr/local/include, so are in either -Isystem or -I path. However, in my code, I want to...

Is sphere.obj attached anywhere here? Not being an actual sphere, but a polygonal representation, I suspect tessellation will affect reproducibility. I think I have reproduced it in Houdini by constructing...

We discussed this in the last TSC meeting, notes can be found at: https://github.com/AcademySoftwareFoundation/openvdb/pull/925/files The problem we suspect is with intentional smoothing we apply to the levelset fields. We do...

> I don't see an ideal solution here as we can't easily modify the constructors due to the tag dispatch class. I think the simplest approach is to just add...

I've added support for union/intersection, though it is somewhat difficult to setup these cases in practice. Two touching cubes tend to convert slightly differently so don't trigger it unless I...

Good point on ambiguity. I think it makes sense to only prune cancelled if pruning is enabled. So I've documented it to that effect & also made it act that...

I've added the SOP layer to match 19.5, if this is okay I'll merge as we are now past 9.1.

Thank you for the heads up! Unfortunately, we are trying to get away from boost, instead trying to move to std:: methods where possible. So stepping backwards for MSVC isn't...