I am learning to use the Qt6 controls with Pyside6, but after the last update I could no longer run this code and it tells me that there is a Segment Fault (‘core’ generated), and before the last update I could run it without problems, now I have tested the same code using PyQt6 and it has not presented any problems and it runs perfectly, I don’t know if you could help me here, and if not I would appreciate it if you could tell me where I can continue searching to solve this error thank you
It is in the screenshot of the error when trying to execute the code
There are some differences between PySide6 and PyQT6. Not many, but sometimes code written for one fails to function in the other.
Both are available and both can be installed at the same time.
Just an FYI
Please always post the text that you copy & paste as preformatted text (the </> button on the toolbar). Images are larger than the equivalent text, are not searchable, and are sometimes difficult to read.
I seem to recall that there has been some changes in pyside6 vs pyqt6. It seems they may now be coming from the same source and merging since there was a lot of code duplicated. Was just reading about that recently.
I’ve encountered the same problem, after updating the entire packages with ‘dnf update’ last week, PySide6, which I am using with venv, spit out a core dump and stopped working, even for a simple ‘Hello world’ type GUI program. I’ve never had this happen before with Fedora.
Since I could not find out which package(s) is/are causing this problem using QT_DEBUG_PLUGINS=1 flag unfortunately, and since I am using some important applications on the PySide6 almost everyday, I had no choice but to switch OS from Fedora 40 to RHEL 9.4 to use these applications with the PySide6.
The actual backtrace for the core dump would probably be useful.
Since you are both working in virtualenvs (and since there is no python-pyside6 package in Fedora 40 anyway*), you are using the binary wheels from PySide6 · PyPI. That doesn’t rule out the problem being in a Fedora package, but it’s relevant.
*At least, not yet – there may be a backport after some packaging issues are worked out in Rawhide.
Are you able to upload a simple test application source that reproduces the problem, similar to the main.py in the first screenshot? I haven’t upgraded my main system to Fedora 40 yet, but it would be interesting to try to reproduce this when I do.