Here's the result:

Let me know if there's anything else I can provide that helps.

I have to say I can't correlate the encoder values you quote with your output or my ParkData file...
Vince

Did a bit more testing today. Load configuration on connect or never makes no difference.

I've also narrowed it does to turning off the mount. I can shut down everything else... Software, pi, computer... It knows where it starts. But if I turn off the mount, it loses its location.

I had a thought... Remember, I put the mount in pc direct mode... I kept turning off the mount from that mode. I tried exiting this mode before turning off the mount... Unfortunately, no difference.

Looking forward to what you can diagnose from the debug log...
Vince

Ok I think I get the point: the driver finds the mount with motors already initialized and thus leaves encoders unchanged even if the mount was parked. I believe that when I wrote the driver, I had in mind that if the mount has been parked, the user should have put off the mount before reconnecting, so the driver should find the motors non initialized. Actually there is no reason to do that. Furthermore it seems that when using the hand controller in direct mode, motors are already initialized before the first connection. I will correct the problem and commit the change asap.

Just commited the change, you can test from the svn repo.

That sounds great... But I'm going to need a bit more guidance on how to test from the svn repo... Not sure what that means.
Sorry for the rookie questions...
Vince