fix: Set toogle button value before connecting it to its slot

[buhtz]

This might fix #1535 .

Problem was that the value/state of the button "Show hidden files" was set after it was connected to its slot. This happened before the main window construction was finished. But it triggered some other events in the main window that couldn't be processed correct because that window itself is not finished with its construction.

Solution: Set the value of that button first and then connect it to its slot.

[buhtz]

Just one line modified. Mhm...
The 1.4.0 release is only some hours old. I vote to delete the "v1.4.0" tag and re-do the 1.4.0 release instead of creating a 1.4.1.

[aryoda]

Problem was that the value/state of the button "Show hidden files" was set after it was connected to its slot.

The same class of problem may still exist for other edge cases unless *all actions are connected at the end of the constructor (__init__) only (or at least when we are sure the GUI state was completely initialized):

backintime/qt/app.py

Lines 129 to 131 in ddd142b

	self._create_actions()
	self._create_menubar()
	self._create_main_toolbar()

We should do some more checks or code reviews here.

The 1.4.0 release is only some hours old. I vote to delete the "v1.4.0" tag and re-do the 1.4.0 release instead of creating a 1.4.1.

I really prefer to not manipulate a published Git history so I would go for 1.4.1.
The efforts are the same.

I can do this since it I am a release prepration Pro now 😉

[aryoda]

@buhtz THX for investigating and fixing this! I could reproduce the problem too now. I will prepare a new minor release Sunday EOD (unless other urgent fixes pop up).

[aryoda]

Ah, the CHANGElog entry is still missing ;-)

[aryoda]

@buhtz FYI: I think we should keep this PR now as it is (no more commits) since it has an impact on the AUR package's checksum (the patch includes new commits):

https://aur.archlinux.org/pkgbase/backintime#comment-934159

[buhtz]

OK. But just for my learning. I don't understand why it has an effect on the AUR because the it should refer to the latest release. So much to learn ... 🤣

[aryoda]

I don't understand why it has an effect on the AUR

The patch in the AUR package references the github PR so the diff changes when we add commits:

https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=backintime#n14