Can confirm! There's supposed to be a button below the text; I guess it went missing! Let me see if I can figure out what happened to it.

Code says this:

                    helpfulAction: {
                        if (BluezQt.Manager.rfkill.state === BluezQt.Rfkill.Unknown) {
                            return null;
                        } else if (btManager.bluetoothBlocked) {
                            return enableBluetoothAction;
                        } else if (root.emptyList) {
                            return addBluetoothDeviceAction;
                        } else {
                            return null;
                        }
                    }

So we're only showing the button when btManager.bluetoothBlocked is true. Probably we're missing another condition in there.

Ok, while debugging this, the button returned! :/ That'll make it more annoying.

Argh, now I can't get it to happen anymore!

Git commit d4da9ece3dc475ccea0bdaec919eb6784604cc5f by Nate Graham.
Committed on 13/06/2024 at 19:56.
Pushed by ngraham into branch 'master'.

applet: remove unnecessary action enablement conditions

These actions are only visible under certain conditions; no need to also
set their enablement statuses to the same conditions.

M  +0    -4    src/applet/package/contents/ui/FullRepresentation.qml

https://invent.kde.org/plasma/bluedevil/-/commit/d4da9ece3dc475ccea0bdaec919eb6784604cc5f

I don't know if that might fix it, but it was also wrong.

(Disclaimer: I do not know anything about neither QML, nor Plasma Applet internals)

I tried to debug things myself, and the preliminary results were pretty weird (like, the `if` branch taken in `text` was seemingly not the same as `if` branch taken in `helpfulAction`), but after some time bug has suddenly disappeared by itself for me too, and I was unable to complete the debugging process.
But, for little I was able to gather before bug has stopped reproducing, it seems to me that the bug somewhere in the QML handling side of things rather than in the QML markup of the applet itself.

I applied your patch, and also added extensive (I hope so) `console.log()` coverage for later:
```qml
                    text: {
                        // We cannot use the adapter count here because that can be zero when
                        // bluetooth is disabled even when there are physical devices
                        console.log("!!! text, btManager =", btManager);
                        console.log("!!! text, root =", root);
                        console.log("!!! text, BluezQt.Manager =", BluezQt.Manager);
                        console.log("!!! text, BluezQt.Manager.rfkill =", BluezQt.Manager.rfkill);
                        if (BluezQt.Manager.rfkill.state === BluezQt.Rfkill.Unknown) {
                            console.log("!!! text 1, rfkill.state =", BluezQt.Manager.rfkill.state);
                            return i18n("No Bluetooth adapters available");
                        } else if (btManager.bluetoothBlocked) {
                            console.log("!!! text 2, btManager.bluetoothBlocked =", btManager.bluetoothBlocked);
                            return i18n("Bluetooth is disabled");
                        } else if (root.emptyList) {
                            console.log("!!! text 3, root.emptyList =", root.emptyList);
                            return i18n("No devices found");
                        } else {
                            console.log("!!! text 4");
                            return "";
                        }
                    }

                    helpfulAction: {
                        console.log("!!! button, btManager =", btManager);
                        console.log("!!! button, root =", root);
                        console.log("!!! button, BluezQt.Manager =", BluezQt.Manager);
                        console.log("!!! button, BluezQt.Manager.rfkill =", BluezQt.Manager.rfkill);
                        if (BluezQt.Manager.rfkill.state === BluezQt.Rfkill.Unknown) {
                            console.log("!!! button 1, rfkill.state =", BluezQt.Manager.rfkill.state);
                            return null;
                        } else if (btManager.bluetoothBlocked) {
                            console.log("!!! button 2, btManager.bluetoothBlocked =", btManager.bluetoothBlocked);
                            return enableBluetoothAction;
                        } else if (root.emptyList) {
                            console.log("!!! button 3, root.emptyList =", root.emptyList);
                            return addBluetoothDeviceAction;
                        } else {
                            console.log("!!! button 4");
                            return null;
                        }
                    }
```
I will try to continue to look into it when the bug will be back again.

Thank you!

Have any of you tried to print rfkill state, both in CLI and from BlueZ point of view?

It happened to me last week that after a whole day of fixing/refactoring Bluetooth applet, next morning my internal Bluetooth device went completely missing, as if it did not exist, and the only clue I had were two lines of cryptic errors in the kernel journal `dmesg`. Luckily, I had an external USB stick so I could continue with my work. I couldn't figure it out though, but after one more day it kinda fixed itself. And that's how I found out that the "Enable" buttons and switches can be out of sync when no devices are present.

Anyone still reproducing this? It's been working reliably for me for a couple of months now with Plasma 6.2 and now current git master(which will be 6.3) and the 6.10 and 6.11 kernels.

🐛🧹 ⚠️ This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information, then set the bug status to REPORTED. If there is no change for at least 30 days, it will be automatically closed as RESOLVED WORKSFORME.

For more information about our bug triaging procedures, please read https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging.

Thank you for helping us make KDE software even better for everyone!

🐛🧹 This bug has been in NEEDSINFO status with no change for at least 30 days. Closing as RESOLVED WORKSFORME.