I’m working on reproducing a piece of wood corner trim. The piece looks like this:
Contouring in just the long direction looks like it will work nicely, but outlining is misbehaving. You can see that the tips of the part are lopped off during outlining:
(I’ve gotten used to “submitting bugs” here on the forum. If you’d prefer that I use the github issue tracker, just say the word. It doesn’t look very current, however.)
It’s easier to keep track of bugs / requests if it’s through the issues tracker, even if that hasn’t been popular of late. You can attach relevant files like the .km file I’m about to ask you for Or just drop it to me in Discord. This is a bit of an interesting case. I know what’s probably going on.
I could make a long-ish video about the complexities of process this shape given the current algorithmic approach. However, I think I will put that energy into the new code base that will take a different approach and handle these situations better. This is a very useful test case, though. So thanks for that.
this undercut is an especially hard case right now because it’s along an edge curved in Z. had the edge been co-linear along Z, it would have been detected. there is a more reliable, but more computationally expensive, way to handle this which I plan to implement in 2.5.
I think that if I make the part wider, and perhaps play with the step down value for the cutter, that I might be able to make the ends that get snipped off be outside the bounds necessary for my part. I already have to do the cutting of the underside myself – it might actually be useful to have “extra” curved surface when I go to trim it (manually) to the right size. (I.e., right now the curve sweeps out 90 degrees. 100 might be perfect, then I’ll cut it to 90.) Sounds good, anyway!