You are not logged in.
I'm on X11, using i3. When I'm talking via Google Meet I am not allowed ("Allow" option being disabled when "Select window or screen" list item is selected in Firefox notification) to share other screens (windows) besides Firefox's own. I've heard that such an issue is common when one is using Wayland, but I don't understand why it exists on X11.
My question is: how can I let or make it possible for Firefox to allow to share other windows on my system?
Offline
about:config, check whether it's set to use xdg-desktop-portal fro screen sharing
https://wiki.archlinux.org/title/Firefo … ntegration
Offline
Which option is that? I've tried setting all widget.use-xdg-desktop-portal.* options to 1 and starting xdg-desktop-portal and xdg-desktop-portal-gtk services, but it helped nothing. I already use Wireplumber and PipeWire. I then tried restarting all the associated services (xdg-desktop-portal, xdg-desktop-portal-gtk, pipewire, pipewire-pulse and wireplumber); still nothing.
Last edited by funeshroom (2025-03-23 15:51:43)
Offline
I've tried setting all widget.use-xdg-desktop-portal.* options to 1
That's pretty much the opposite of what you want.
Offline
Alright, I tried also setting them all to 0 and nothing changed.
Last edited by funeshroom (2025-03-23 15:51:59)
Offline
generally bugged session? How are you starting i3 and what's your output for
loginctl session-status
printenv DBUS_SESSION_BUS_ADDRESS Offline
sanity check: FF is from the repos, not some flatschpak?
Try to run it from the console and see whether there're error messages explaining this restriction.
Offline
sanity check: FF is from the repos, not some flatschpak?
Try to run it from the console and see whether there're error messages explaining this restriction.
FF is from the repos.
If I run FF from the console it does not write any output to it, and also logging to a file with MOZ_LOG_FILE environment variable just produces an empty file. I've tried both separately and together while attempting to share another window.
On the other hand, I've figured that sharing another window works fine if that window is on the same workspace as the Firefox instance is; however, I have to select another window from the notification list item menu itself, meaning I can't click on it ("Select window or screen" list item is still disabled).
Last edited by funeshroom (2025-03-22 10:10:05)
Offline
generally bugged session? How are you starting i3 and what's your output for
loginctl session-status printenv DBUS_SESSION_BUS_ADDRESS
I'm using LightDM display manager (with the GTK+ greeter) to start i3.
Here is the output of loginctl session-status, it's not much:
9 - funeshroom (1000)
Since: Sat 2025-03-22 10:52:12 CET; 2min 6s ago
State: active
Leader: 5554 (lightdm)
Seat: seat0; vc7
TTY: tty7
Remote: no
Service: lightdm
Type: x11
Class: user
Desktop: i3
Idle: no
Unit: session-9.scope
├─5554 lightdm --session-child 12 20
├─5568 i3
├─5584 transmission-daemon
├─5594 light-locker --no-late-locking --no-lock-on-suspend --no-lock-on-lid
├─5595 xss-lock --transfer-sleep-lock /home/funeshroom/.local/share/xss-lock-transfer-sleep-lock-i3lock.sh
├─5596 autocutsel -s PRIMARY
├─5597 autocutsel -s SECONDARY
├─5598 autocutsel -s CLIPBOARD
├─5599 /bin/sh /home/funeshroom/.local/bin/xselsave
├─5601 dunst
├─5602 i3bar --bar_id=bar-0 --socket=/run/user/1000/i3/ipc-socket.5568
├─5621 i3status
├─5677 /usr/lib/firefox/firefox
├─5750 /usr/lib/firefox/firefox -contentproc -parentBuildID 20250311232648 -prefsHandle 0 -prefsLen 36928 -prefMapHandle 1 -prefMapSize 264964 -sandboxReporter 2 -chrootClient 3 -ipcHandle 4 -initialChannelId {477cb769-2774-4f02-99cf-6f7cd17edc56} -parentPid 5677 -crashReporter 5 -appDir /usr/lib/firefox/browser 1 socket
├─5768 /usr/lib/firefox/firefox -contentproc -isForBrowser -prefsHandle 0 -prefsLen 37031 -prefMapHandle 1 -prefMapSize 264964 -jsInitHandle 2 -jsInitLen 247456 -parentBuildID 20250311232648 -sandboxReporter 3 -chrootClient 4 -ipcHandle 5 -initialChannelId {ffecdbc1-e83c-40bd-95a5-7ac5148b5ce3} -parentPid 5677 -crashReporter 6 -greomni /usr/lib/firefox/omni.ja -appomni /usr/lib/firefox/browser/omni.ja -appDir /usr/lib/firefox/browser 2 tab
├─5772 /usr/lib/firefox/firefox -contentproc -parentBuildID 20250311232648 -prefsHandle 0 -prefsLen 37031 -prefMapHandle 1 -prefMapSize 264964 -sandboxReporter 2 -chrootClient 3 -ipcHandle 4 -initialChannelId {3ca90a5f-6e0c-4865-b296-d98cdd001def} -parentPid 5677 -crashReporter 5 -appDir /usr/lib/firefox/browser 3 rdd
├─5826 /usr/lib/firefox/firefox -contentproc -isForBrowser -prefsHandle 0 -prefsLen 44918 -prefMapHandle 1 -prefMapSize 264964 -jsInitHandle 2 -jsInitLen 247456 -parentBuildID 20250311232648 -sandboxReporter 3 -chrootClient 4 -ipcHandle 5 -initialChannelId {17f21a58-a52b-4ede-ad19-e571fa0da4f7} -parentPid 5677 -crashReporter 6 -greomni /usr/lib/firefox/omni.ja -appomni /usr/lib/firefox/browser/omni.ja -appDir /usr/lib/firefox/browser 4 tab
├─5880 /usr/lib/firefox/firefox -contentproc -parentBuildID 20250311232648 -sandboxingKind 0 -prefsHandle 0 -prefsLen 44995 -prefMapHandle 1 -prefMapSize 264964 -sandboxReporter 2 -chrootClient 3 -ipcHandle 4 -initialChannelId {3acdc5e1-7cc3-4446-91f1-4c65cb157062} -parentPid 5677 -crashReporter 5 -appDir /usr/lib/firefox/browser 5 utility
├─5895 /usr/lib/firefox/firefox -contentproc -isForBrowser -prefsHandle 0 -prefsLen 40542 -prefMapHandle 1 -prefMapSize 264964 -jsInitHandle 2 -jsInitLen 247456 -parentBuildID 20250311232648 -sandboxReporter 3 -chrootClient 4 -ipcHandle 5 -initialChannelId {9da32c97-5471-4b41-9dde-d5a7cf600896} -parentPid 5677 -crashReporter 6 -greomni /usr/lib/firefox/omni.ja -appomni /usr/lib/firefox/browser/omni.ja -appDir /usr/lib/firefox/browser 6 tab
├─5898 /usr/lib/firefox/firefox -contentproc -isForBrowser -prefsHandle 0 -prefsLen 40542 -prefMapHandle 1 -prefMapSize 264964 -jsInitHandle 2 -jsInitLen 247456 -parentBuildID 20250311232648 -sandboxReporter 3 -chrootClient 4 -ipcHandle 5 -initialChannelId {6fbb3df2-4ead-462e-8daa-af7f5af7ddcb} -parentPid 5677 -crashReporter 6 -greomni /usr/lib/firefox/omni.ja -appomni /usr/lib/firefox/browser/omni.ja -appDir /usr/lib/firefox/browser 7 tab
├─5918 /usr/lib/firefox/firefox -contentproc -isForBrowser -prefsHandle 0 -prefsLen 40542 -prefMapHandle 1 -prefMapSize 264964 -jsInitHandle 2 -jsInitLen 247456 -parentBuildID 20250311232648 -sandboxReporter 3 -chrootClient 4 -ipcHandle 5 -initialChannelId {ab4099b2-c353-4719-8068-04b41c2bf9a5} -parentPid 5677 -crashReporter 6 -greomni /usr/lib/firefox/omni.ja -appomni /usr/lib/firefox/browser/omni.ja -appDir /usr/lib/firefox/browser 8 tab
├─6002 /usr/lib/firefox/firefox -contentproc -isForBrowser -prefsHandle 0 -prefsLen 40542 -prefMapHandle 1 -prefMapSize 264964 -jsInitHandle 2 -jsInitLen 247456 -parentBuildID 20250311232648 -sandboxReporter 3 -chrootClient 4 -ipcHandle 5 -initialChannelId {c8fc4357-f3c4-4f65-824c-9c75fa6d7710} -parentPid 5677 -crashReporter 6 -greomni /usr/lib/firefox/omni.ja -appomni /usr/lib/firefox/browser/omni.ja -appDir /usr/lib/firefox/browser 9 tab
├─6034 xsel -kn
├─6062 alacritty
├─6070 /usr/bin/bash
├─6077 loginctl --no-pager session-status
└─6078 xsel -i
Mar 22 10:52:12 archlinux systemd[1]: Started Session 9 of User funeshroom.
Mar 22 10:52:13 archlinux transmission-daemon[5584]: net.cc:307 Couldn't bind port 51413 on 0.0.0.0: Address already in use (98) -- Is another copy of Transmission already running? (net.cc:307)
Mar 22 10:52:13 archlinux transmission-daemon[5584]: net.cc:307 Couldn't bind port 51413 on ::: Address already in use (98) -- Is another copy of Transmission already running? (net.cc:307)The output of DBUS_SESSION_BUS_ADDRESS:
unix:path=/run/user/1000/busOffline
Addendum: When I share a window from the FF's workspace and then I move it to another workspace, everything continues to work as it should — the window is being shared just as if it had remained on the FF's workspace. Still nothing is written to the logs or to the console.
Last edited by funeshroom (2025-03-23 15:52:59)
Offline
If I run FF from the console it does not write any output to it
Did you kill all previous intances?
Because otherwise FF will just IPC-open a new window and exit.
I've figured that sharing another window works fine if that window is on the same workspace as the Firefox instance is
Please test the behavior on an openbox session - it basically sounds as if FF just ignores unmapped windows (for whatever reason) so let's see whether specific to the i3 implementaiton of "workspaces" (virtual desktops)
Offline
Yes, I did kill all previous FF instances (that's what I hadn't done at first though, which is why I've edited my post later and after trying the same but now on the first FF instance).
Last edited by funeshroom (2025-03-23 15:53:14)
Offline
On an Openbox session the behavior is very similar: I can still only select windows from the current workspace (and in the same way i.e. from the main notification menu, meaning I can't click on them), but now if I move the shared window to another workspace, it is replaced by a black screen in the FF stream, and moving it back to the workspace where the FF window is (or vice versa) restores it in the stream.
Last edited by funeshroom (2025-03-23 15:53:25)
Offline
Please don't bump, but edit your post if nobody has yet replied and also avoifd bloating the thread with pointless full quotes.
I guess you're running a compositor (picom) along i3, what prevents the windows from being unmapped?
Overall sounds like it's just a FF UI bug that's artifically constraining the windows you get to share - there's an open bug for at least windows itr
https://bugzilla.mozilla.org/buglist.cg … +selection
=> Same/similar problem w/ chromium?
Offline
No problem, I'm sorry.
I am not running any compositor.
I've tried with Chromium and the behavior is exactly the same as on FF. But when I'm using Chromium, when I switch from a workspace where Chromium is not to the one where it is, windows opened on the workspace I switched from are visible for a brief period in the Chromium's screen share selection window before they are replaced by the windows on the current workspace.
Offline
You're not going to be able to share unmapped windows - they technically don't exist (hence the black content)
You stand a better chance doing this "across workspaces" w/ a wayland compositor or a compositing WM that doesn't actually unmap the windows (to keep the redirected for live previews in the tabbox or whatever)
Interestingly i3 doesn't seem to do that, but maybe only stacks them behind its container window?
Did you test chromium on i3 or on openbox?
Offline