<GitHub>
[solvespace] whitequark commented on issue #292: @KmolYuan The "solver" branch exists because of copyright reasons, not for technical ones. If you are aiming for integrating your Python bindings, that should be done in the "master" branch, where active development happens. Are you fine with that? https://github.com/solvespace/solvespace/issues/292#issuecomment-333271204
<GitHub>
[solvespace] jwesthues commented on issue #295: Commercial CAD systems generally do a fixed chord tolerance, either absolute (like for export) or as a function of model size and/or zoom level and stuff. They presumably have lots of special case heuristics on top of that. If you've never seen the PWLs there, then I'd guess it's because they're just faster, making better use of the GPU, selectively re-meshing, etc. So they can set a finer tolerance w
<GitHub>
[solvespace] jwesthues commented on issue #295: Add a "chord angle" tolerance, on the maximum angle between the curve and the chord approximating it? So if you want big things meshed finer, then you make the chord distance tolerance smaller, but if you want little things meshed finer then you make the angle finer. That's how SolidWorks export does it, for example. https://github.com/solvespace/solvespace/issues/295#issuecomment-333275504
<GitHub>
[solvespace] jwesthues commented on issue #295: And easy to implement--just modify `MakePwlWorker` to compute some tangents and check the angle in the same place it checks distance. To make the Booleans consider angle tolerance would be a bigger project; but that's relevant only in cases where the chord distance tolerance is so big that the Booleans are failing entirely, and you can always fix that by dropping the chord distance tolerance, same as