Hello Gents,
Some head-scratching for me, maybe for you too.
I have tested the your model Russ, it's fun and super promising. And then I figured I might test a dummy version of something with over hanged and reversed faces.
The results are quite interesting.
1. A lateral hole drilling does not seem to apply the tool path inside the hole for some reason. It's clearly the correct depth or orientation, but not the right position.
2. Interestingly, I feel like a "minus" somewhere in the code would solve all angle positioning issues
3. The depths in the OP seem to not affect the outcome (final depth) and the finishing depth breaks the algorithm.
I give you my file for your personal fun.
One issue that I cannot seem to repeat is that upon asking for the lateral hole negative it machined the blind side of the hole instead of the hole side of the hole. But I suspect my statement #2 is the underlying reason...
This is fun! A lot of fun! Thanks