Hey MacOS users!  Please do some testing to see if the latest beta version is working well.  We are looking at doing a release of 3.5.8 soon and would like to verify that everything is working well.  I'm going to try directing you to the new Nightly builds instead of building a Mac DMG this time.  That way you can let me know if there are any issues with the nightly builds or with the latest version of KStars for the Mac.

binary-factory.kde.org/view/MacOS/job/KStars_Nightly_macos/

Thanks,

Rob

Hey Rob,

thank you very much for the nightly builds for Mac.
I would love to test the nightly but somehow I already spend hours getting Baader Steeldrive II to work on m Mac without success: indilib.org/forum/focusers-filter-wheels...t-connect.html#81157

Oliver

Hi Rob,

thanks for the nightly builds! 
I have tested the 1462 build yesterday on Big Sur. The first thing I noticed is that it does not start using dark theme colors, I had to change the theme manually and solved it.
Also it crashed several times in the middle of the session. As for the rest it's all working ok, I think
The attached zip has logs on the two objects I tried.
 

Yeah the theme colors change is one across the board.

github.com/KDE/kstars/commit/06d6c046058...6d937910570541af989f

I much prefer the dark theme, but it got changed by this person.  In reading the commit message, I see there was a reason for it, but I still prefer the dark theme.

I found an issue with the Meridian Flip option:

I selected a target in KStars, and slewed to it.  Then I used the alignment tab to "Slew to Target" and it worked great.  It indicated that the "target" was within acceptable range.  So obviously a target was set.

But when I tried to set the Meridian Flip parameters, it says that the status is inactive because "no target was set".

I went back and tried again, by selecting a target with the mouse, and whose name then was properly displayed in the upper left corner.  I then realigned using Slew to Target - and that completed perfectly.  But again - the Meridian Flip section says "no target set".

I tried again by choosing a different target in the "Choose Object" menu, and while I had no trouble slewing and aligning on it, the Meridian Flip system still says no target.

I then quit this 3.5.8 beta version of KStars, opened up the latest 3.5.7 Stable version and could not reproduce the error.  All as fine, with a perfect meridian flip.

Peter


 

Thanks Rob !
I've tested KStars 358 only with the simulator and found no problem at all.

Jean-Claud, did you try using the Meridian Flip settings? I just tried with the simulator and just like real life, it says NO TARGET was set. even though one clearly was.

Jean-Claud, did you try using the Meridian Flip settings? I just tried with the simulator and just like real life, it says NO TARGET was set. even though one clearly was.

 

Hi Peter,

The Meridian Flip/No Target bug was actually addressed by Jasem pretty recently (yesterday I think).  The nightly builds only happen once a day, so it is possible that the next time it builds the error will be gone.  Check it out tomorrow.  Also note, I just moved some build code from the craft recipe into cmakelists.txt. It works fine on my computer, but I will have to wait until the nightly build tomorrow to find out if it works on the KDE Binary Server.  So if that doesn't work you might have to wait another day to see if the Meridian Flip bug is fixed.

Thanks,

Rob

OK Peter, Rob gave the answer !

Looks like we will have to wait at least one more day, I forgot the prefix path in one of my copy commands.  It is now updated so hopefully it will work tomorrow.   Crossing my fingers.

Hi Rob,

Great work on getting nightly builds working. I'm on MacOS 11.6 (BigSur) and I tried the most recent nightly build. I run INDI on a RPi and Kstars on my Mac.
- It fired up fine and loads my saved config fine.
- It connected to all my equipment as expected (2 ZWO cameras, ZWO EAF, ZWO EFW, EQmod Mount, Pegasus PPBA).
- EKOS polar alignment worked as expected.
- Slewing the scope from Kstars worked as expected (as did Kstars itself).
- Internal guider calibration and guiding worked. I did notice, however, that the guide S/N would regularly drop to zero before returning to its previous value. There was no obvious reason for this (no clouds, dew, etc.). I see others have mentioned this on the forum with 3.5.7 so its probably not specifically 3.5.8 (my guess). I haven't noticed this myself with 3.5.7 though. I'll include a screenshot from Analyse where this is very obvious.
- Focusing (Linear Model) worked as expected.
- I setup a sequence on M101 and scheduled it.
- I watched the schedule for a couple of hours and everything worked as expected.
- The schedule continued for a few hours before capture ran into some problems (could have been cloud, not sure). Looks like guiding was also having problems so maybe the sequence was aborted and restarted as it looks like Focus then got into a loop where it tried every 3 mins or so and failed each time. After 3+ hours of focus fails, Kstars crashed. I didn't have logging on as I find this causes me more crashes than when its off, so no more info I'm afraid.
- Took flats this morning. All worked as expected.

I haven't seen repeated Focus fails like this before so maybe its a scheduler change to restart slightly differently from 3.5.7 and maybe cloud was causing the problem? So generally most things worked as expected and the problems may not be specific to 3.5.8

I would like to build 3.5.8 myself. In the past I've got  github.com/rlancaste/kstars-on-osx-craft  working but when I just tried again I'm getting errors. Before I invest effort in getting this working is this still the way to go (I see you've been putting lots of effort into infrastructure work like StellarSolver so I'm wondering if this will have affected the build process)?

Once again, thanks for all the work you're doing!