I'm using the OSX version of FET 5.35 provided by Darren McDonald.
When I close FET without saving modifications, FET asks me "Your data file has been modified - do you want to save it?".
When I select "No" in that dialog, FET crashes, presenting the regular OSX error dialog.
Can you please sent me the log file?
Senden Sie mir bitte den "Bericht" per Email (siehe www.timetabling.de )
Sadly i can't reproduce that bug.
Does it also happen with "empty" files. So you just open fet, maybe add a subject and close again withoutout saving?
Or does it happen "only" with your dataset? In that case please sent me your dataset per email or let me know how i can reproduce the bug. please don't forget to sent me the logfile (Bericht).
It's easy to reproduce here. Opening FET without loading anything (I guess this is what you mean with empty dataset), adding something, closing without saving -> crash.
Thank you for your report!
It is easy to follow it, as it is an assertion failed, not a segmentation fault.
I cannot replicate it on my GNU/Linux. I'll try now on Windows and let you know.
If you look at src/fetmainform.cpp (FET sources), line 999, you'll see that the code is right. I don't know what should I do to solve this problem. It is maybe a Mac OS X interface bug or Qt.
Edit: I tried on Windows 10, no crash.
Very stange, since i compiled it myself under MacOS and it doesn't crash. (but i compiled with Qt 5.9, maybe a new bug in Qt 5.10?).
Quote from: Liviu Lalescu on January 16, 2018, 07:57:22 PM
If you look at src/fetmainform.cpp (FET sources), line 999, you'll see that the code is right. I don't know what should I do to solve this problem. It is maybe a Mac OS X interface bug or Qt.
If that's really the crash point.
@math, do you press Cancel button? Just close the dialog? Press the No button?
In FET I press CMD-Q for quitting the application.
Then FET asks if I want to save to modifications
Now I press the "no" button.
I tried on my GNU/Linux and Windows 10, everything is OK (I cannot replicate your bug report).