I've mentioned previously my weirdness with the PA routine here: indilib.org/forum/general/374-ekos-plate...limitstart=0&start=6. I'm getting used to my workaround at the moment, but I admit that because of it (in particular, my extra plate solve/sync), I am unsure of the validity of the error value being reported.

I've found a couple of other things that are minor issues:
* When doiing the alt error checking, the result is always opposite to whatever it says. If it says x too low, then I have to lower the mount, not raise it.
* It always comes in with Northern hemisphere set for the Az testing - could it use the site information to determine the correct hemisphere for the user, i.e. Southern hemisphere for me?

You are absolutely right regarding the hemisphere selection, there shouldn't be any it should be determined from the user site. Regarding the direction, it should be opposite to northern hemisphere, so let me double check simulation-wise and hopefully tonight with the actual telescope as well, IF the sandstorm decides to subside today

Excellent. And I hope the sandstorm passes soon - at least that's one thing I don't have to deal with!

It did and I just tested it less than an hour ago, the results reported are consistent. It reported 2:20 too low, and I raised a bit, and re-measured again and it reported 0:40 too low which I was happy with. Same with azimuth error, it reported 1:10 too east, I used to az knob to move it west, and recalculated and it is now 0:15 too east. But this is how I define north/south, west/east here in the northern hemisphere:

1. North: the mount altitude INCREASES. The object appears to go UP.
2. South: the object appears to go DOWN.
3. West: The object appears to go LEFT.
4. East: The object appears to go RIGHT

I'm looking at this via the polar finder scope. I will test southern hemisphere under simulation later on and will see how it fairs.

Here is just a frame I took in the last 5 mins!

Nice!

Back on the PA stuff. I get confused with the East/West thing anyway, so I move it one way and see what happens - if it decreases I was doing it correctly, otherwise go the other way.

However the too low/high thing is based on the alt degree value - for me, if it says to low, I have to make the alt degree value decrease. Too high, increase the alt degree value.

Re: hemisphere selection - you might want to override it if you're just plate solving... but that's very much an edge case you can easily handle by using the plate solver separately to ekos.

For the alt testing choice of east/west, could it also determine that from the mount's position?

Also, are you going to use the user's site or the telescope's site, or are they effectively the same thing in the program by this point? (I note they can be different if you don't use kstars to update the telescope when it connects).

The KStars site location is used for pretty much everything. I just removed the "east/west" direction selection as well, since Ekos can determine this from the telescope coords. Now, regarding the low/high, I do it the opposite way, if it was too high, I lower the mount altitude by adjusting the altitude knob, polaris would appear to go DOWN in the polar finder scope. I'm not sure what star do you use in the south, but the principle is the same.

If kstars uses it's geographic site for everything (makes sense), can we get an option to have it update its site from the telescope? I know it can update the telescope's site, but I'd like the other way round. I have a couple of sites configured in the scope and would like to have the software adjust as required.

And/or add site management to kstars where I can preset a number of sites and switch between them.

You can select this from "Update Source" in KStars Options --> INDI. Just set it to "Device" and it will use the geographic location and time from the device.

Any chance for a mixed mode? Time from computer to mount; Location from mount to computer?

Also, I've just been trying the PA routine again for Alt. It's definitely reporting the reverse of what I expect for high vs low.
A question so I can determine if the work-around process I'm doing is valid:
The GOTO mode is set to sync for these tests - it's the only way to get the workaround to "succeed"
I start the Alt error check and it plate solves.
It then moves 30' in RA.
At this point if I leave it alone, nothing happens.
I update the mount details in the options, switch to GOTO mode and do a capture.
This succeeds and of course syncs the mount.
After the sync, the 2nd PA image is taken and solves and reports an error.

My concern is the middle plate solve/sync is messing up the results - could these be affecting the calculations used? Maybe the reason I'm getting things in "reverse" could be because of this?

Plus, thanks for the auto-detection of hemisphere, etc - very nice.