You are not logged in.
Hi,
running a daily updated Arch, Wayland, GNOME, terminator as my terminal and fish shell. I use the alt-gr+dot key combination a lot, for the last command's argument. This has been working since - I believe - yesterday, maybe since 2 days ago, not sure. Now whatever I try, the AltGr (my right Alt key) key works as a no-op: running eg. `cat` and pressing alt-gr+<any key>, it shows the key literal.
Trying to hunt down the issue with `xev`, alt-gr+dot shows:
KeyPress event, serial 38, synthetic NO, window 0x3000001,
root 0x4f9, subw 0x0, time 10242599, (76,113), root:(276,391),
state 0x80, keycode 0 (keysym 0x1ff, abovedot), same_screen YES,
XLookupString gives 0 bytes:
XmbLookupString gives 2 bytes: (cb 99) "˙"
XFilterEvent returns: False
and DeepSeek tells me the "abovedot" part shows that it's a "dead key". I verified the issue running `weston` to avoid any GNOME-related setting playing games with me, and it also happens there.
Any ideas what could have wrong here in the last 2 days?
Last edited by kjozsa (2025-01-16 16:22:35)
Offline
Some additional debugging info:
$ setxkbmap -query
WARNING: Running setxkbmap against an Xwayland server
rules: evdev
model: pc105
layout: us
variant: intl
$ xmodmap -pm
xmodmap: up to 3 keys per modifier, (keycodes in parentheses):
shift Shift_L (0x32), Shift_R (0x3e)
lock Caps_Lock (0x42)
control Control_L (0x25), Control_R (0x69)
mod1 Alt_L (0x40), Alt_L (0xcc), Meta_L (0xcd)
mod2 Num_Lock (0x4d)
mod3 ISO_Level5_Shift (0xcb), Hyper_L (0xcf)
mod4 Super_L (0x85), Super_R (0x86), Super_L (0xce)
mod5 ISO_Level3_Shift (0x5c)
Offline
For weston and gnome (likely on wayland) xev and setxkmap aren't authorative - let alone xmodmap.
=> You'd ideally test the behavior on X11
Regardless, on X11 xev shoudl™ show altgr (by itself) as ISO_Level3_Shift, if that's not happening check whether "evtest" and "libinput debug-events" register the physical key at all.
Any ideas what could have wrong here in the last 2 days?
Do you have a spare keyboard?
Local humidity changes like from a radiator or a wildfire or so?
(Depending on the keyboard hardware this can lead to static build-up and block the electrical signal in the keyboard)
Offline
@seth this is a 1 month old HP ZBook Firefly laptop, I would score out any hardware related issues. I digged through the BIOS but I also think it's set up correctly.
Now I tested the problem throughoutly in the evening, and I'm getting sure it's a GNOME issue. Previously:
- I claimed the issue exists on weston - not true, the issue only exists on weston started within my gnome session from a terminal (it starts in a nested window)
- starting from gdm, neither weston nor hyprland has the problem, alt-gr+dot works flawlessly in terminals there. Also started from gdm, GNOME on X.org does have the issue though.
At this point I believe that my terminal setup, my fish shell and Wayland does not cause any trouble, but GNOME does.. and I still believe it's a recent issue introduced in the last few days.
Reviewing the upgraded packages in `pacman.log` from the last day, I suspect maybe one these might have caused it:
[2025-01-15T09:55:56+0100] [ALPM] upgraded util-linux-libs (2.40.3-1 -> 2.40.4-1)
[2025-01-15T09:55:57+0100] [ALPM] upgraded systemd-libs (257.2-1 -> 257.2-2)
[2025-01-15T09:55:57+0100] [ALPM] upgraded brltty (6.7-2 -> 6.7-3)
[2025-01-15T09:55:57+0100] [ALPM] upgraded cryptsetup (2.7.5-1 -> 2.7.5-2)
[2025-01-15T09:55:57+0100] [ALPM] upgraded docker-compose (2.32.2-1 -> 2.32.3-1)
[2025-01-15T09:55:57+0100] [ALPM] upgraded intel-gmmlib (22.5.4-1 -> 22.5.5-1)
[2025-01-15T09:55:57+0100] [ALPM] upgraded linux-firmware-whence (20241210.b00a7f7e-1 -> 20250109.7673dffd-1)
[2025-01-15T09:55:57+0100] [ALPM] upgraded linux-firmware (20241210.b00a7f7e-1 -> 20250109.7673dffd-1)
[2025-01-15T09:55:57+0100] [ALPM] upgraded python-anyio (4.7.0-2 -> 4.8.0-1)
[2025-01-15T09:55:57+0100] [ALPM] upgraded util-linux (2.40.3-1 -> 2.40.4-1)
[2025-01-15T09:55:58+0100] [ALPM] upgraded systemd (257.2-1 -> 257.2-2)
[2025-01-15T09:55:59+0100] [ALPM] upgraded systemd-sysvcompat (257.2-1 -> 257.2-2)
[2025-01-15T10:01:29+0100] [ALPM] upgraded google-chrome (131.0.6778.264-1 -> 132.0.6834.83-1)
[2025-01-15T10:01:29+0100] [ALPM] upgraded firefox-gnome-theme (133.1-1 -> 134-1)
[2025-01-15T11:06:35+0100] [ALPM] upgraded firefox (134.0-1 -> 134.0.1-1)
[2025-01-15T11:06:35+0100] [ALPM] upgraded git (2.48.0-1 -> 2.48.1-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded polkit (125-1 -> 126-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded gnome-color-manager (3.36.0+r74+g19b6f67d-1 -> 3.36.2-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded gnome-software (47.3-1 -> 47.4-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded gupnp-av (0.14.1-2 -> 0.14.3-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded libwnck3 (43.1-1 -> 43.2-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded ollama (0.5.4-3 -> 0.5.5-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded qt6-webengine (6.8.1-1 -> 6.8.1-2)
[2025-01-15T11:06:36+0100] [ALPM] upgraded rsync (3.3.0-2 -> 3.4.0-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded tree-sitter (0.24.6-1 -> 0.24.7-1)
[2025-01-15T11:06:36+0100] [ALPM] upgraded xdg-user-dirs-gtk (0.11-3 -> 0.12-1)
[2025-01-15T11:08:20+0100] [ALPM] upgraded aws-cli-v2 (2.22.34-1 -> 2.22.35-1)
Offline
The only thing w/ a remote impact on the situation should be systemd[-libs] and possibly util-linux[-libs], but it would be very odd if they only impacted gnome.
On gnome/X11, what's the actual complete "xev -event keyboard" ouput when trying to enter altgr+˙ ?
Offline
it looks like this (press + release):
KeyPress event, serial 28, synthetic NO, window 0x2200001,
root 0x4f9, subw 0x2200002, time 4732827, (67,46), root:(605,481),
state 0x0, keycode 108 (keysym 0xfe03, ISO_Level3_Shift), same_screen YES,
XKeysymToKeycode returns keycode: 92
XLookupString gives 0 bytes:
XmbLookupString gives 0 bytes:
XFilterEvent returns: False
KeyPress event, serial 28, synthetic NO, window 0x2200001,
root 0x4f9, subw 0x2200002, time 4732995, (67,46), root:(605,481),
state 0x80, keycode 60 (keysym 0x2e, period), same_screen YES,
XLookupString gives 1 bytes: (2e) "."
XmbLookupString gives 1 bytes: (2e) "."
XFilterEvent returns: False
KeyRelease event, serial 28, synthetic NO, window 0x2200001,
root 0x4f9, subw 0x2200002, time 4733097, (67,46), root:(605,481),
state 0x80, keycode 60 (keysym 0x2e, period), same_screen YES,
XLookupString gives 1 bytes: (2e) "."
XFilterEvent returns: False
KeyRelease event, serial 28, synthetic NO, window 0x2200001,
root 0x4f9, subw 0x2200002, time 4733219, (67,46), root:(605,481),
state 0x80, keycode 108 (keysym 0xfe03, ISO_Level3_Shift), same_screen YES,
XKeysymToKeycode returns keycode: 92
XLookupString gives 0 bytes:
XFilterEvent returns: False
Offline
AltGr+. is perfectly recognized but apparently there's no 3rd level defined in the current layout
setxkbmap -print -query # on X11 this will actually tell what's currently used
Offline
$ setxkbmap -print -query
WARNING: Running setxkbmap against an Xwayland server
xkb_keymap {
xkb_keycodes { include "evdev+aliases(qwerty)" };
xkb_types { include "complete" };
xkb_compat { include "complete" };
xkb_symbols { include "pc+us+inet(evdev)" };
xkb_geometry { include "pc(pc105)" };
};
rules: evdev
model: pc105
layout: us
The strange part is that I get the exact same output when logged on to Hyprland from GDM, and it works there in the terminal..
Offline
That's not from an X11 session?
Try to use the "intl" variant, the basic us layout doesn't use the 3rd level shift
Offline
No, that all happens on Wayland (but I had the same result from GNOME over X11). I switched to the intl variant but same effect, only a dot char gets printed to terminals.
Now it looks like this:
$ setxkbmap -print -query
WARNING: Running setxkbmap against an Xwayland server
xkb_keymap {
xkb_keycodes { include "evdev+aliases(qwerty)" };
xkb_types { include "complete" };
xkb_compat { include "complete" };
xkb_symbols { include "pc+us(intl)+inet(evdev)+level3(ralt_switch)" };
xkb_geometry { include "pc(pc105)" };
};
rules: evdev
model: pc105
layout: us
variant: intl
options: lv3:ralt_switch
Offline
Please try this on an X11 session - setxkbmap isn't authorative nor can control the keyboard layout in a gnome/wayland session.
That aside, does the xev output change? Does the input behave differently in xterm (ie. an actual xwayland client)?
Offline
I had an idea and created a brand new local user, logged on to it (using GNOME on Wayland) and alt-gr+dot works there also. That means that must be some local setting in my user's home...
Note that `setxkbmap` output looks exactly the same for the new user and my own:
setxkbmap -print -query
WARNING: Running setxkbmap against an Xwayland server
xkb_keymap {
xkb_keycodes { include "evdev+aliases(qwerty)" };
xkb_types { include "complete" };
xkb_compat { include "complete" };
xkb_symbols { include "pc+us+inet(evdev)" };
xkb_geometry { include "pc(pc105)" };
};
rules: evdev
model: pc105
layout: us
I'm a bit lost on how to find the difference between the two user's local settings though.
Offline
DeepSeek helped to solve it: `dconf reset -f /org/gnome/desktop/input-sources/`, and relogin - now it's all good.. *pheeew*. Thanks much @seth for trying to help me out on that one, appreciated!
Last edited by kjozsa (2025-01-16 16:25:24)
Offline
Input method configuration?
Again: setxkbmap and an empty bag on wayland are worth the bag. It doesn't mean a lot and especially not for native wayland clients.
The relevant configuration is the gnome settings, https://help.gnome.org/users/gnome-help … ts.html.en
Edit: F5
Last edited by seth (2025-01-16 16:34:52)
Offline