Thanks for doing the testing!

That means that whatever caused your problem, it was introduced pretty recently.

It has been introduced between 3.5.1 and 3.5.2.
I thought you did the changes to solve the ZWO issue ?

So the only issue with ZWO was that the single library file that you copied into 3.5.1 was missing so that ZWO didn’t work. I didn’t really do anything to fix the problem, I just made sure to delete the broken link from my build folder when I built 3.5.2.

So I was saying that if you confirmed that 3.5.1 guides fine with the library copied into the app, then whatever is causing your issue must have been introduced between those 2 versions. It should not be StellarSolver or ZWO driver related since those 2 things didn’t change between those versions, so it must be a change somebody made to the guiding code in KStars in the last couple of weeks.

Please tell me how I could help !

At the moment I am not sure since I haven't made any changes since 3.5.1 myself. I will try to let some people know and we will see what they say.

Jean Claude,

I'm not sure what to say. This is what the log shows: In 3.5.2 after a single 1000ms pulse you move out a lot in RA, more than 15 pixels which is enough to terminate the calibration in that direction. When the calibration then tries to go back, by presumably pulsing in the other direction (the logs don't explicitly show the pulses, you could add debug mount logging for both ekos and indi to show that) the mount continues to move out in RA (not back). It keeps pulsing 1000ms to come back, but the mount keeps going out. Eventually calibration gives up. So, this is not a stellarsolver thing.

The calibration logic is doing the right thing. The star detection seems reasonable. However, for some unknown reason, the mount is not responding to the pulses to come back, or the pulses are not being sent as they should. I don't think the calibration logic was changed in 3.5.2. I'm a bit stumped, if it is a software problem. If not for the success of the 3.5.1 code, I would have guessed that there was something wrong with your hardware (e.g. unbalanced so the reverse RA pulses are ineffective).

This business of a single pulse completing the RA-outward calibration is unusual.

Can you compile and run source code? I could make a version of 3.5.2 with more diagnostics. Unfortunately, I'm no longer up-to-speed with compiling on the Mac (I could help if you also had a raspberry pi or linux computer environment).

Hy

On Linux the popup looks good. On Mac, the popup is shown with the same width as on Linux but the buttons and text don’t scale well. Wait, I’ll take a few screenshots.

Here's a screenshot of the popup on Linux. This is how the popup is shows without resizing it:



Here on Mac without resizing



and with resizing



Wouter

Hy & Rob,
I'm sorry to say that I'm unable to compile and run source code.
This is not a hardware problem because previous KStars versions run OK on my setup.
I'm astonished that nobody else reported this bug.

Perhaps not many people have tested this beta version as thoroughly as you. I, for one, have not tested guiding, mainly because my mounts don't work very well at the moment, and will see if I can get the calibration done, because that part does work on Linux. It's just that the grease of the mounts seem to have dried out of have become very dirty and needs to be replaced, which I am working on. However, I use ASI120MC cameras for guiding so I don't know if that makes a difference. Is it possible for you to test your gear with KStars nightly on Linux, even for only one night?

I use only macOs and not Linux. Sorry.
But you said that the calibration does work for you on Linux.
This is precisely the calibration which can't be completed on macOS (hence no guiding is possible).
This might be a leading thread to find the bug.

Jean-Claude,

Unfortunately, the issue here is not as simple as all guiding/calibration doesn’t work. I tested calibration and guiding on both my rig and the simulator, and it worked just fine. There are many folks that have tested versions of 3.5.2 without noting this issue. There is something different about your environment triggering the problem. I’m happy to continue to try to help debug this. Jasem just released 3.5.2 on all platforms, so if it’s an issue more general than your particular environment we should see some more reports soon.

Could you please generate one more log with those two mount debugging options turned on? Please also fully describe your gear, for example focal length of your scope, guide scope or OAG, mount, and cameras. Could you also try to reproduce the issue with the simulator?

I also private messaged you, please respond if you want to look at this together over Zoom or Google Meet.

Thanks
Hy