Well, I'm on a row today... Anyways...

I've been using qobi's fix for the Su-8w for a while now, with no problems, with exception of a nasty tendency to, after a few keyboard timeout disconnects (the keyboard won't remain turned on after a few minutes, I assume it auto-shuts down in order to save battery), no longer respond to Ctrl-other keys combinations. I noticed that after a few disconnects, the keyboard, both internal and external ones, would no longer be able to output those Ctrl commands, as in... Ctrl-backspace would no longer evoke task switch, ctrl-shift-X would no longer evoke Xterm, etc, until I rebooted the device. I was actually planning on creating a new thread on this once it bugged me enough.

Well, today I attempted to find some workaround for this problem, so I wouldn't have to reboot the device and close all open MicroB tabs. Once ctrl commands became unresponsive, I tried unpairing the devices, and re-pairing them. To my surprise, keyboard and cell phone would pair up alright, although the bluetooth light on the Su-8w kept blinking, much in the same way as back when main.conf was disabling HID profiles!!

I tried rebooting, then. The keyboard would be recognized by the N900, but woudn't create any response whatsoever. Typing random characters wouldn't summon contacts, as usually, neither key combos would work. I tried summoning Xterm through the internal keyboard, and typing "setxkbmap -device 4 -I -I/usr/share/X11/xkb-chinook -rules base -model nokiasu8w -layout us", as I normally would in order to set the layout for the keyboard, and then, it would work, although only up until next disconnect, when, even upon pairing, the keyboard would remain unresponsive, instead of only with the wrong layout, as usually.

Ok... So I have a functional bluetooth keyboard, albeit one that won't work unless it has its layout set in xterm, as opposed to normal behaviour, when I would normally be able to summon Xterm through the keyboard, layout as incorrect as it were, set the correct layout, and go about my business... And a random bug which I know nothing about, and that I can't fix.

So... Any ideas on what's going on? I'd really like to be able to revert to proper behaviour.

Did you try extkbd?

Not yet. I should have, but I never got around to uninstalling the custom su-8w custom layout file for proper testing. Has extkbd exhibited this behaviour?

Still, I'm now afraid that I have a "dirty" testbed, as whatever the issue in my device is, it might be carried over to extkbd testing... Or is that not the case?

I probably should test it anyhow, considering I'll likely need to tinker with the whole bluetooth setting.

Maybe you are right, the issue seems to be with bluetooth stack, which is out of control of extkbd. Anyway where did you get custom su-8w layout. I may include it in some future version of extkbd.

Any ideas on how to explore your guess that the problem lies with the bluetooth stack? I'm not afraid to research on the subject, but being a linux newbie means I'm not even sure where to start looking for a solution, so any help, either through pointing me in the right direction or taking me through a solution will be greatly appreciated.

As for the custom layout, it's a portuguese, mainly PT-BR oriented one. Alt-Gr C produces cedillas, " produces dead diaresis, while Alt-Gr " produces normal quotations, and the varios diacritics are changed to dead ones.

Let me know if you still want it, I'll link it here.

Ok, so... Out of the blue (ironic), the keyboard has resumed its normal behaviour.

I noticed that, after the third or forth timeout disconnection in the abnormal state, the keyboard wouldn't pair with the N900 automatically, in contrast to reported behaviour of pairing up but presenting no reaction to keystrokes before setting the layout. Not pairing up correctly, as in, bluetooth blue light flashing on the keyboard, as in can't pair with the device.

Ok, so, I turned bluetooth off and back on. Then turned on the keyboard, and summoned busybox. The keyboard once again paired up automatically, albeit with no input on screen. Once I gave the command to set the layout I got this:

Goes without saying, but no keyboard input. Afterwards, I turned bluetooth off and on again, turned on the keyboard, and it paired up just fine, again, still with no input. Summoned busybox again, pressed the up arrow key tp recall the above command... And it worked fine, much to my surprise. Keystrokes still wouldn't register in the device without settting the layout, but still, the error message presented above did not reappear.

The truly good news came with the next reboot, which I had to do due to a particular MicroB misbehaviour (after a while, all new tabs lock up in eternal loading until the device is rebooted). Once the device went through startup, I pressed a key on the keyboard by chance, and the keystroke summoned contacts... That came as a surprise.

Can anyone make sense out of this? Well, seems to problem fixed itself by recreating the conditions which first caused it.