> After turn on the controller and connect to kstars, the mount already
> unparked, I do not have to press unpark.

I don't use Park/Unpark since I don't have an observatory, so I am not sure
if what you describe has always been the case for OnStep, or something new
in OnStepX

> The position always set to NCP (though the scope is parking horizontally).

This tells me that something is not right.
Ask in the OnStep mailing list if this is how things are supposed to work or
there is an issue somewhere (whether a bug in the firmware, or the steps
of your workflow).

Hi,

I just updated OnStep to version: = 10.19.c since it removes a bug (onstep.groups.io/g/main/message/55149)
and this update also changes the behavior after cold start (what you call power cycle) .

Now when I am parked and shutdown OnStep I am still in that state when restart.
I need to unpark and then I work like usual.

Thank, Howard, welcome to the forum. It's great to see you have your account finally sorted out.
I am using 10.17a I think I'll update to the latest firmware and see if it solves my problem.
@Alain: I'll upgrade to 10.19c and see what happens. Also will test new driver when I have the chance. Thanks!
@Khalid: as a first timer on OnStep, I find the learning curve is quite steep, but I am trying to catching up. Thanks for your help!

@all,

After update to 10.19.c,
changing some cosmetics in the indi User Interface,
fixing the type about reading calculated errors from OnStep

it seems to work fine.
OnStep Alignement values are kept even after cold reboot of OnStep (Power Off > Power On)
Parking status is kept as well enven after cold reboot.
Now it is as I was used to,
when restarting I find the mount in Idle status and position is in the park position.
Then I unpark and voilà! OnStep unpark and tracking is enabled.

The changes are already merged into master.

Now I go back to my hot potatoe, the observatory

Thanks Alain, that is great. Just what I need.
Btw the dome looks cool. The shell almost see through, what is it made of? And what is the rolling mechanism?

Have a look on: github.com/azwing/Observatory/tree/main/doc
and pictures on azconcept.org/albums-photo/wppaspec/oc1/lnfr/cv0/ab6


it's a bit outdated since I changed the the cylinder against the dome but the principles remain the same
Structure is wood, roof is polyester and glas fiber

Okay so after contacting Howard I got my mount to park and work as expected. Turns out I was falling into a bug that's on the OnStepX, I was parking too low and the scope pointed to the East while parking West and that's what confused the OnStep parking mechanism, Howard fixed that in 10.19.f and it's now parking correctly.
But I ran into another problem, The mount just won't do Meridian flip correctly when running Scheduler on Kstars. When Approaching the meridian, the mount just stops tracking and then INDI keeps outputting (Goto in progress) when I try to initiate a goto.
My setting is as follows:
1. Automatic Meridian Flip at Limit: (on SWS) : Off
2. Meridian Flip, Pause at Home (on SWS): Off
3. Meridian Limit (on SWS): West : 1 ; East: 1 (I am not sure if I should put 1 or -1 into the limit?)
4. Meridian Limit (on mount panel in EKOS): 3 deg
Im I missing something here?

1. Automatic Meridian Flip at Limit: (on SWS) : Off

I think that might be the cause of your problem. It is the mount that decides to and performs the meridian flip. KStars just issues a goto that the mount decides requires a flip.

Thanks, Ed.
How should I put the meridian limit on the mount and on Ekos? For my CEM70, if I set meridian limit on my mount to say 7 degs then I have to set it to lower than 7 on Ekos and the mounts perform just fine. Is OnstepX's settings the same?

I believe it's similar but you'll have to test with your kit to determine effective limits.

Okay so here's how the setting should be with OnStep for it to work with Scheduler automation
I put this here for reference for anyone else run into the same problem as I did:
1. The flip limit on Mount panel in Ekos should be set to 1-2 deg pass meridian.
2. The flip limit on OnStep should be set to a number higher than what you have on Ekos (if Ekos has 1 deg then OnStep should have 4-5 deg, the number should be positivel
3. The preferred pier side should be EAST (this is important, without this, the mount wont flip).
That's all. The mount should now flip in tandem with Ekos scheduler.

I have ran into this weird little bug on ekos and OnStepX (10.19.k) i am not sure it's an OnStepX problem or a problem with how the INDI driver work. The problem is as following:
1. Action set to Sync in the Alignment tab.
2. After a goto and plate solve, there is an amount of pointing error. (Lets say it's RA:250 " DEC: 300"). The INDI perform a sync with the scope and it says the sync is completed successfully.
3. Now if the mount is tracking and you continue to perform a platesolve without any goto/slew then you'd expect the new platesolve error to be near Zero (since the scope pointing was synced with the last platsolve) But no, after the solve, the error still showns exactly as before. (Still RA:250 " DEC: 300"). This will continue no matter how many time you do the platesolve/sync. The only way to bring the error to zero is by changing the action to "Plate solve and slew". The mount will platesolve, sync the error, then goto the new position, after the second sync, the goto error now is corrected.
Is this how the sync should work on OnstepX or is it a bug? I think because of this, the mount unable to build a model correctly with KStars?