Added a number of features including self-calibration

[thereza]

So I got sick of trying to guesstimate the size of the box to calculate the mm per pixel numbers. I wrote some code to move the camera head and measure how much the homing dot moves to determine the x/y mm per pixel. The results are highly repeatable so it looks good to me. I then took it a step further and measured the amount of slack in the system along the 4 directions of movement. If you move from left to right, then move left again the same distance, you will not end up in the same spot due to the slack in the system. I measure this difference with the goal of being able to take the slack into account for all movements rather than relying on the slack compensation mechanism currently in place. Here are values I get in my system. The two numbers represent movement from left to right and right to left (for x) and up to down and down to up (for y). These numbers are also highly reproducible (within .01mm). However, I'm not sure what the source of the slack is -- I tightened my x belt significantly and it had zero influence on the x slack value. The measurements and calibration only take a few seconds total.

slack x : 0.3092785 / 0.3070458
slack y : 0.1862599 / 0.1786845

I would like to submit the enhancements back but I have no clue how git works and it just looks like a mess to use (I'm a fan of SVN). I'll try pulling the latest version, merging the changes I want, and pushing that back out.

Reza

[WayOutWest]

However, I'm not sure what the source of the slack is -- I tightened my x belt significantly and it had zero influence on the x slack value.

Did you ever find out the source of the slack?