Buck

Yes that is the case that needs to handled differently for what you want. I don't think anything else needs to change but I haven't given it much thought if...

Thanks for the bug report. I made some changes in the Rust code to fix issues around overlapping cases, but it looks like this case maybe still has some problems...

OK looks like the issue (at least for the Rust code) is the fuzzy comparison epsilon value used in the circle-circle intersect function is too small/not matching the position equal...

In addition to changing the intersect epsilon values I also needed to make the polyline segment intersect function be more consistent across cases - specifically for the line-arc intersect case...

Note the commit above wasn't implemented quite right, I fixed a bug introduced by it with this commit: https://github.com/jbuckmccready/cavalier_contours/commit/1a40c08a2a23f18969e9b77c2976e244fea473d1 (see commit changes for details).

The stitch selector functions are used to resolve the case where there are multiple slices that could be stitched together at the same point in 2D space. `currSliceIndex` is used...

1. I think that should work... 2. I do not think there is an easy way to do this right now without modifying the code (code would need to change...

Seems like an interesting thing to solve with arcs for performance/fidelity. I don't know when I'll get around to this - and if I do then it will be in...

The algorithm always produces offset curves which are exactly parallel to the input curve (unless the offset is too large and it becomes collapsed). To maintain exactly `2` units away...

It's not supported/implemented, but there is an algorithm to efficiently do it. I'm in the process of rewriting this library in Rust. See my replies to this issue: https://github.com/jbuckmccready/CavalierContours/issues/29