<p>
On Wed, Sep 09 2020, Adam Nielsen wrote:<br />
</p>
<p>
>> I have remapped<br />
>> `CapsLock` to `Backspace`<br />
>> `AltGr` to `ControlR`<br />
>> `ControlR` to `AltR`,<br />
>><br />
>> Now if i type one of the next sequences, it will be well interpreted :<br />
>> • `ControlR` + `AltR` + `Backspace`<br />
>> • `ControlR` + `AltL` + `Backspace`<br />
>> • `ControlL` + `AltR` + `Backspace`<br />
>><br />
>> However if I type this one :<br />
>> • `AltL` + `ControlR` + `Backspace`<br />
>><br />
>> xev will not see any input.<br />
>> This one definitivly make me think it’s the firmware fault.<br />
><br />
> If I understand correctly, you are saying that ControlR + AltL works,<br />
> but the reverse (AltL + ControlR) does not. It could be a firmware<br />
> issue however it’s also possible it can be a keymap issue.<br />
</p>
<p>
The position of the keys matters in this issue, to be more clear,<br />
another example :<br />
if I use `setxkbmap -model pc104 -model us -option’, so there is no<br />
options, a clean and regular keymap and Alt<sub>R</sub> is <b>not</b> an<br />
ISO<sub>Level3</sub><sub>Shift</sub> but a regular Mod1 like Alt<sub>L</sub><br />
`xev’ will not see any input with the following combinaison :<br />
`Alt<sub>L</sub>` + `Alt<sub>R</sub>` + `Caps<sub>Lock</sub>`<br />
However, it will see the input of<br />
`Control<sub>L</sub>` + ’Control<sub>R</sub>` + `Caps<sub>Lock</sub>`<br />
</p>
<p>
It is indeed dumb because no one need to press the two Alt together,<br />
but it demonstrate the issue : theses particuliar keys (no matter what<br />
keysym is attributed) doesn’t respond when pressed together and it is<br />
finally, not related with remapping.<br />
</p>
<p>
> If you want to know for sure if it is a keyboard/firmware problem,<br />
> change the mapping to some other keys. Instead of mapping AltGr to<br />
> ControlR, map say ShiftR to ControlR instead. If you try a few<br />
> different key mappings and you still have the same problem, it is more<br />
> likely to be your layout at fault rather than the hardware.<br />
</p>
<p>
Yes, my tests leads me all the time to the combinaisons with the same<br />
keyscodes, no matters what keysyms are attributed. What really confuse<br />
me is it is I do not understand the logic behind it, some keys accept more<br />
than others and always others specific combinaisons.<br />
</p>
<p>
> The combinations you have - two modifiers + letter - should work on<br />
> pretty much any keyboard. If you were using a non-modifier key then<br />
> you could run into issues after holding 2-3 keys depending on the<br />
> keyboard design.<br />
><br />
> You can also look into tools that dump Linux evdev events. This would<br />
> allow you to see the key press / key release events at the kernel<br />
> level. A firmware problem would mean that at some point pressing a key<br />
> won’t generate an evdev event, but if you are seeing evdev events but<br />
> no Xorg key event, then it means it’s definitely an issue with your Xorg<br />
> keymap.<br />
</p>
<p>
If `xev` a sufficient tool for this task I think I am done and can<br />
conclude that my keybaord is not terrible nor good.<br />
</p>
<p>
Thank you Adam Nielsen and Dave Howorth fos your insights which where<br />
more valuables that the computer constructor.<br />
</p>
<p>
–<br />
</p>