Hi Ken and Jared,
I was setting up and testing notifications using the simulators in daytime. I intentionally triggered recovery mode by enabling star fading in PHD2. When recovery started, there was a temperature-triggered Auto Focus run. Since the camera simulator in SGP was producing a cloud-like image, AF kept getting bad HFR and moving the focuser out. This went on repeatedly until I manually canceled the AF. (I got 52 notifications stating that AF got a bad HFR 3 times in a row, so it did not look like it was ever going to stop.)
It seems like this could happen in an actual unattended session. If AF runs and clouds are present, could AF run forever, continually moving the focuser out? Would it make sense for there to be a limit of outward AF motion before aborting the sequence, like two or three times the number AF data points? I think the “endless” AF run could cause problems with tracking past the meridian, running past sequence end time, etc. Should AF failure behave like other failures during recovery (failure to center, failure to settle guider, etc.) ?