×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

KStars 3.6 crashes when Find Object is selected

  • Posts: 200
  • Thank you received: 28
Thanks Gilles - it's happening to me whether I use Control-F (in my case Command-F) or the magnifying glass. Prior to 3.6 everything worked.
1 year 7 months ago #85077

Please Log in or Create an account to join the conversation.

  • Posts: 401
  • Thank you received: 41
Confirm the issue. Crush as soon as I select the magnifying glass, on M1 Mac with 3.6 stable.
Last edit: 1 year 7 months ago by Euripides.
1 year 7 months ago #85079

Please Log in or Create an account to join the conversation.

  • Posts: 294
  • Thank you received: 54
Just installed KStars 3.6 stable on my i386 Macbook Pro (2013) and the problem is not present on this architecture.
1 year 7 months ago #85080

Please Log in or Create an account to join the conversation.

  • Posts: 200
  • Thank you received: 28
When I run a screen sharing session from my Mac to the RPi (which is also running 3.6) Find Object works just fine.

I have a 2015 vintage Intel Mac that I just installed 3.6 on. Most of the time Find Object crashes - 5 out of 7 times. The only times it worked were after I had tried Find Object from within the Observation Planner. If I have a successful Find Object in the Observation Planner and then exit the Observation Planner, I can execute a Find My from the main menu without a crash.

From within the Observation Planner, Find Object works about 3 out of 4 times but still crashes so it's not quite bullet proof either.

So bottom line, my Intel Mac is somewhat more reliable than my M1 Mac but neither is performing a Find Object in a stable manner.
1 year 7 months ago #85082

Please Log in or Create an account to join the conversation.

  • Posts: 200
  • Thank you received: 28
Hi Gilles - a quick question - what version of macOS are you running on that MacBook Pro. My wife has a model of that vintage and macOS 12.x won't run on it.
1 year 7 months ago #85083

Please Log in or Create an account to join the conversation.

  • Posts: 460
  • Thank you received: 69
I get the same thing with Kstars 3.6.0 stable on Monterey 12.5 on a MacStudio.
Thread 1:: com.apple.rosetta.exceptionserver
0   runtime                       	    0x7ff7ffc2a944 0x7ff7ffc26000 + 18756
1   runtime                       	    0x7ff7ffc383e0 0x7ff7ffc26000 + 74720
2   runtime                       	    0x7ff7ffc39e58 0x7ff7ffc26000 + 81496
 
Thread 2:: com.apple.CFSocket.private
0   ???                           	    0x7ff8a4bfe940 ???
1   libsystem_kernel.dylib        	    0x7ff814734d5a __select + 10
2   CoreFoundation                	    0x7ff8148562db __CFSocketManager + 643
3   libsystem_pthread.dylib       	    0x7ff8147694e1 _pthread_start + 125
4   libsystem_pthread.dylib       	    0x7ff814764f6b thread_start + 15
 
Thread 3:: QDBusConnectionManager
0   ???                           	    0x7ff8a4bfe940 ???
1   libsystem_kernel.dylib        	    0x7ff8147330aa poll + 10
2   QtCore                        	       0x113dba00e qt_safe_poll(pollfd*, unsigned int, timespec const*) + 94
3   QtCore                        	       0x113dbb85c QEventDispatcherUNIX::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) + 812
4   QtCore                        	       0x113d58d17 QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) + 471
5   QtCore                        	       0x113b972fc QThread::exec() + 140
6   QtDBus                        	       0x10fd3bf93 QDBusConnectionManager::run() + 19
7   QtCore                        	       0x113b9820a QThreadPrivate::start(void*) + 330
8   libsystem_pthread.dylib       	    0x7ff8147694e1 _pthread_start + 125
9   libsystem_pthread.dylib       	    0x7ff814764f6b thread_start + 15
etc.
1 year 7 months ago #85085

Please Log in or Create an account to join the conversation.

  • Posts: 593
  • Thank you received: 276
I too have this issue. Unfortunately when I run it under debug it works fine :S

A way around this is to navigate in the sky and click on the object you want (obviously avoiding the find object dialog box). Its a bit tedious but does work.
1 year 7 months ago #85104

Please Log in or Create an account to join the conversation.

  • Posts: 294
  • Thank you received: 54
Hi Fred, I run macOS 11.6.8 on my MacBook Pro.
1 year 7 months ago #85105

Please Log in or Create an account to join the conversation.

  • Posts: 401
  • Thank you received: 41
So we can replicate the issue on M1 processors for sure. Probably @Rob could take a look on that.
1 year 7 months ago #85156

Please Log in or Create an account to join the conversation.

  • Posts: 84
  • Thank you received: 6
I have the same problem on Stellarmate/RPi, when I try to use the search function in kStars or in the scheduler, kStars crash and restarts, It’s 32bit Stellarmate 1.7.2, kStars 3.6.0.
Last edit: 1 year 7 months ago by Keld Henningsen.
1 year 7 months ago #85300

Please Log in or Create an account to join the conversation.

  • Posts: 200
  • Thank you received: 28
Not sure if I should continue to post here or file a bug report but I'm attaching a cut and paste of the log file that Apple generates when an application crashes. Hoping to image this weekend and will try to use a work around of some sort so that KStars will run.

Clear Skies,
1 year 7 months ago #85301
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 2876
  • Thank you received: 809
Hey guys, after some very challenging testing involving a gradual process of elimination this week, I think I have tracked this issue down. It seems to have been caused by this commit:

github.com/KDE/kstars/commit/5a2ba9f8e8b...93a50ca66f09cb2f985d

It was quite hard to track down because any time I launched Kstars from the terminal or from QT Creator on my system, the crash didn't occur, but it did occur whenever I ran it by double clicking it in the latest release. That made it very hard to debug. I had to add a bunch of qMessageBox messages to even see where it was crashing and what was causing it. I think it was caused by something in the constructor for the new Asynch DB that was introduced in this commit.

I found that if I put my QMessageBox in the right spot in the constructor, the crash didn't occur on my Mac. So then I tried replacing that with a sleep command and it seemed to work too. So it isn't a really good solution, but it is a band aid until this can get resolved in a better way. Hopefully it works for everyone.

I just built a new DMG today that should include this change and I uploaded it to the normal download spot for 3.6.0. Please test to see if it is fixed.
The following user(s) said Thank You: Frederick Ruegsegger
1 year 7 months ago #85539

Please Log in or Create an account to join the conversation.

Time to create page: 0.345 seconds