Get really bizarre results in this beta when using subtract circle or cylindrical solid.

Hi Kevin, thanks for reporting this bug - I get the same result in the prior Mar-24 V3 beta and in V2 as well so it does not seem to be a newly introduced bug at least...

It seems to have something to do with the cylinder seam edge just barely grazing right along the bottom plane of the other object. If you rotate the cylinder by 90 degrees so the seam is in some other spot like the attached version it should then work ok.

I'll put this on my list of things to investigate but since it's not a recent regression it's not likely something that I'd be able to fix for v3 since messing around with intersection mechanisms is a sensitive area and it's quite easy to introduce problems in other cases when tweaking things.

- Michael

Hi Kevin, it will also work without rotating it if the circle's center point is repositioned to be more exactly right on the bottom edge. Currently it's located something like 0.0013 units just very slightly above it, things that are barely grazing each other at just right around 0.001 units or so (which is the intersection tolerance) can be somewhat problematic in general.

In the attached version I've moved the cylinder and circle downwards to true them up so their centers are right on the same level as the bottom plane, this then should also work ok.

Generally what happens with this kind of barely grazing thing is that there's some teeny tiny little intersection piece generated and it's collapsed or back tracking in such a way to mess up the orientation of the trimming boundaries. If a trimming boundary has any self-intersections or backtracking little pieces in it, it makes it not have a well defined inside and outside region anymore and that's when you can see that weird "leakage" looking stuff.

- Michael

Ok, thanks for the explanation, its good to know how to solve it when modelling stuff and getting odd results.