The Astropheric forecast, which shows smoke and smoke forecast (based on wind patterns) showed that the Kincaid wildfire smoke might clear out the second half of the night. So I went ahead and set up, figuring I'd just toss the first half of the night's data. Since the moon is down, I swapped back to my color camera, my ZWO ASI294MC Pro, with my newly-acquired Astronomik CLS-CCD light pollution filter. I previous had the regular version of the CLS filter, but I learned that it doesn't block UV/IR, which was fine with my DSLR because it has its own blocking filter, but the 294 does not. This caused the brighter stars to have weird rings around them, since the IR light doesn't get focused the same way the optical light does in a refractor.
The images that came down were pretty much garbage, as expected. Darn wildfires! Stupid climate change!
I went out and re-focused before bed because the temperature had dropped quite a bit, and then went to bed. In the morning, I looked outside to check on it as I usually do, and it wasn't pointing at the home position! I looked at the sequence to see what happened, and it was still trying to recover, which was weird. It was on my last target, the Flaming Star Nebula, and it turns out that when I copied the sequence to make the new one for the color camera, the "end sequence at" time didn't copy. Whoops! So I set that and parked the mount and buttoned everything back up.
M31 still had tracking issues, even though I made sure it didn't start until 25 minutes after the meridian. That should have been enough time for it to settle down. The Flaming Star Nebula, also west of the meridian, did not have these issues. More troubleshooting to do...