Which coordinate changed?
Two fixes around my Mini Delta printer concerned coordinates, but at different layers. One was in my calibration script. The other was in the firmware’s decision about which mesh points needed extrapolation.
My script measured an adjustment, then sent it as an absolute mesh value. The correction changed the parameter passed to M421:
-self.M421(I=I, J=J, Z=offset+0.001)
+self.M421(I=I, J=J, Q=offset+0.001)Z set the mesh value; Q adjusted it. The accompanying message changed from saying the mesh was set to the offset to saying the selected point was adjusted by it. The patch also restored the gauge position and updated the script’s saved movement state.
The small 0.001 term was a separate workaround. The discussion reported that an exact zero could be treated as an unmeasured point and overwritten by extrapolation. Changing the coordinate mode did not solve that problem.
In the firmware report, the X contribution to a radius calculation was squared, while the variable called ysq held an unsquared signed displacement. That could make the radial calculation invalid or classify a valid probe point incorrectly. The patch supplied the missing sq() call.
The maintainer confirmed bugs in that area and reported the fix in release 1.3.3. My PR closed without merging.
The printer profile records the broader machine configuration. These defects lived closer to the calibration loop. Naming a variable offset or ysq suggested a meaning, but neither name made the operation use that meaning correctly.