AU

`.param` is in radians AFAICT. Probably it should be marked as private (`_parmas`). ``` angles = [0, 45, 0, -90, 0] rotX = [] for angle in angles: # changing...

What I can see is that the last solve in your example is not fully converged (`assy.solve(5)` to get some diagnostics). If you rerun the last solve, it does return...

If you don't like the generated orientation you can always use an aux spine to define one explicitly.

I suppose you are looking for this: ```python from numpy import linspace, sin, cos, vstack, pi from cadquery.func import * from cadquery.vis import show # params t1 = 0 t2...

You could use `apply` and `cq.Wire.positionAt`, e.g. ```python N = 12 res = ( cq.Workplane() .box(1,1,1) .wires('>Z') .apply(lambda x: list(x[0].positionAt(p/N) for p in range(N))) .sphere(0.1, combine='cut') ) ``` 

Is that what you want: `result.objects['m'].loc` ?

AFAICT there is no well defined location of the tagged object. You'd need to construct it yourself from its center and some prior knowledge or assumptions and then multiply with...

I guess you want something like this then? ```python name = 'b2' tag = 'v' loc = result.objects['b2'].loc tagged = result.objects['b2'].obj._getTagged('v').val() result: Vector = tagged.moved(loc).Center() ```

Confirmed, FWIW a workaround: ```python import cadquery as cq from cadquery.func import wire, ellipse from math import pi eps = 1e-6 s = cq.Sketch().face(wire(ellipse(1,1).trim(0,2*pi-eps)).wires().offset(-0.1, mode='s') ```

Very interesting @lorenzncode ! I vaguely remember having similar issues in a container and for me the solution was to choose a different SW rasterizer (`llvmpipe` vs `softpipe` but not...