The theme overrides were not working, so the issue is likely in how the Qt5 template layer was built/integrated in the universe MATE build (Qt6ct seems unaffected, but will not help with code built on Qt5). i.e. the applications may not crash, but will look garbled and exhibit undefined GUI behaviors on legacy desktops. Overriding the themes and other documented workarounds will only suppress standard warnings, but have not had any impact on the core Qt5 issues.
For an LTS OS intended to be around for 12 years, I am not sure things have turned out well in user space and Qt5.
Hopefully, this saves someone a few hours of chasing this external issue in their own projects. =3
Uhm, I use Mate under Trisquel 11 (bound to Ubuntu 22 releases, but libre), it's the default desktop DE and I have no issues with neither qt5ct nor GTK theme overriding, which is more apt for Gnome3/OSTree based desktops than normal setups.
This issue appeared in Canonical's Ubuntu 24.04.1 Noble Numbat MATE Desktop iso and universe repository meta packages ubuntu-mate-desktop and mate-desktop .
The meta package ubuntucinnamon-desktop does not seem to impact the Qt5 Apps/installation in a detrimental way.
If I could offer any advice about this mystery issue... probably wiser to defer upgrading for a bit =3
...why? I know this is the default opinion, but it's always seemed misplaced to me, even more so in 2024.
QT is run by a for-profit company ($QTCOM, TIL!) selling commercial licenses, Electron is maintained by OpenJS, who has it MIT-licensed all the way. QT uses a language that was designed in 1985 (literally 1-5y after GUIs first appeared) and Electron uses a language designed in 1995 (for the express purpose of modernized GUIs). QT stands alone on its monolithic rock, whereas Electron leverages Chromium and Node.js, two absolute powerhouses of free development and dependencies. Finally, and most importantly: the web is more beautiful and far more consistent than native-styled apps, and GUIs made whole-cloth from QT are almost always too ugly to even be in the running there. All of those downsides are worth it to save some RAM? Not even worth it, but worth lauding like it's a brave stance?
Sorry, just triggered my trauma from having to work in QT before I was able to find my true calling as a webslinger. No offense to the author of this particular app ofc, I'm sure it was the right choice for them and it looks well-executed for QT. ...Though if they used Electron, it could've been "Jocker" or "Tocker" (ts!) or "Chrocker" or "Electrocker" rather than "cock-er", but that's neither here nor there.
The best / cheap solution for great wifi is to purchase additional cheap ones, disable the dhcp server and drag lan from one to the other. (AP mode I guess.)
Lounge, Office, Bedroom for example. Connect your devices to each at least once and then forget about it as it'l switch over.
Of course if you have money, you could do the same but using MESH/Mimo.
> Connect your devices to each at least once and then forget about it as it'l switch over.
Not by default, you'd still need to tweak the Roaming Aggressiveness/Sensitivity. Set it too low and it will only switch when the current WiFi is down, too high and it will switch too often and you'll have too many network interruptions.
On iOS devices for example I think the roam trigger threshold is not user selectable (–70 dBm, and if gaining 8/12dBm depending on transmitting state). So setting this up reliably is closer to luck than science.
There are wifi extenders that should do transparent roaming quite well. Some can be connected using RJ45, some just take the wifi and "repeat it" further away, Some can do both.
In this case, you don't even need to connect to each.
As the article mentions in section 17, WiFi extenders are not recommended unless you don't mind slashing your speed in half. Two repeaters will slash your speed to a quarter, three to an eighth, etc. While they do extend range, they do so effectively by halving the speed because they are repeaters, which repeat every data packet they receive to get it to the other side, which wastes half of the available bandwidth. If it is only using RJ45, and not repeating the WiFi signals, then it may be OK but at that point it's acting more as an AP than a repeater. Section 18 of the article suggests an AP (eg. router in AP mode) is the best way to extend WiFi range.
I've had bad experiences with those repeaters, but never used the RJ45 backed ones. IMHO, that is then the same as an AP though - except for maybe sharing the same network SSID.
