Vision Crashes When Measuring Circles

It would appear as though I have multiple problems with this program. I will do my best to explain, hopefully that will help understand.

I have a cylindrical part.
The load instructions state to align a feature upright in a rotary table, and you'll be able to run the program. This part is easy.
However, with other parts I've done, I could find a feature to align upright, and it would work regardless of it was at 0° or 180°, because the features were the same on both sides.

This part is different.
I have a hole at one location at 0°, but at 180° there is no hole. So I'm trying to use "OnError" to determine if the hole is there. If it's not, rotate 180° and try again. If it is, continue the program.

I was able to get this section to work. I set the OnError command to Edge_Not_Detected, GoTo (Label that contains the other circle), which just rotates 180° and measures the same circle.

A variable is created, determining if I am using the circle at 0° or 180° in relation to the rotab (so I know if I need to rotate and re-align or not)

The rest of the program continues when the hole is actually measured.

This worked fine and well, so I moved on to the rest of the program (there are only 48 crossholes of varying diameters and locations, not a waste of time at all!)

When I finished the program (without creating dimensions, I just wanted to see if it would even run before I messed around with them!) that section stopped working.

Not only that, every Vision circle I tried to measure in that program crashes PC-DMIS!

So I more or less created a [theoretically] good program, but it doesn't want to run through for some reason.
----------------------------------------------------------------------------------------------------------------------------------------------------
Randomly throughout the program, Everything would change and be funky.

Vectors of {0,0,1} were transforming to random gibberish (sometimes they would even be in the same plane I left them!)
More often than not 0,0,1 would become {-.000000015648, .0946521, -.99999784135484} or some such nonsense (notice the sign change)
XYZ values would jumble around a few thousandths intermittently
Fun!

Well, I read that pretty much to prevent that, switch all the {0,0,1} with {0+0,0+0,1}, so I went through the painstaking process of adjusting all of them.
I didn't really understand, because the ones I set to 0+0 would stay 0, but ones (say, X 2.765) would change. So I changed stuff to more or less 2.765+0.

But it doesn't work. I can measure holes with Vision in other programs, I even added some line measurements to this one right after the probe change but before the holes. They work fine! The holes don't!

It would appear as though I have multiple problems with this program. I will do my best to explain, hopefully that will help understand.

I have a cylindrical part.
The load instructions state to align a feature upright in a rotary table, and you'll be able to run the program. This part is easy.
However, with other parts I've done, I could find a feature to align upright, and it would work regardless of it was at 0° or 180°, because the features were the same on both sides.

This part is different.
I have a hole at one location at 0°, but at 180° there is no hole. So I'm trying to use "OnError" to determine if the hole is there. If it's not, rotate 180° and try again. If it is, continue the program.

I was able to get this section to work. I set the OnError command to Edge_Not_Detected, GoTo (Label that contains the other circle), which just rotates 180° and measures the same circle.

A variable is created, determining if I am using the circle at 0° or 180° in relation to the rotab (so I know if I need to rotate and re-align or not)

The rest of the program continues when the hole is actually measured.

This worked fine and well, so I moved on to the rest of the program (there are only 48 crossholes of varying diameters and locations, not a waste of time at all!)

When I finished the program (without creating dimensions, I just wanted to see if it would even run before I messed around with them!) that section stopped working.

Not only that, every Vision circle I tried to measure in that program crashes PC-DMIS!

So I more or less created a [theoretically] good program, but it doesn't want to run through for some reason.
----------------------------------------------------------------------------------------------------------------------------------------------------
Randomly throughout the program, Everything would change and be funky.

Vectors of {0,0,1} were transforming to random gibberish (sometimes they would even be in the same plane I left them!)
More often than not 0,0,1 would become {-.000000015648, .0946521, -.99999784135484} or some such nonsense (notice the sign change)
XYZ values would jumble around a few thousandths intermittently
Fun!

Well, I read that pretty much to prevent that, switch all the {0,0,1} with {0+0,0+0,1}, so I went through the painstaking process of adjusting all of them.
I didn't really understand, because the ones I set to 0+0 would stay 0, but ones (say, X 2.765) would change. So I changed stuff to more or less 2.765+0.

But it doesn't work. I can measure holes with Vision in other programs, I even added some line measurements to this one right after the probe change but before the holes. They work fine! The holes don't!