It was only an example, but sure that may have done it.
The only point is … you needed portal for a dependency … and gnome was suggested because its first in the list.
You could probably see this in your logs like
I remember to have learned few months ago XFCE was not 1:1 alternative of GNOME, rather it provides an alternative to GNOME on special way in which it is (in some extent) based on GNOME- reason why set of GNOME user apps/utilities were in use here. That might be context in which GNOME was added to this Manjaro very well.
No it isnt.
As stated above…
You installed something that required desktop portal as a dependency … and gnome was suggested because it is first in the list alphabetically.
You may be sorta referring to how XFCE and Gnome are both GTK desktops? And/or conflating it with the ‘forks of classic gnome’ that is Mate, etc ?
How bad is it then to have ‘gnome-desktop’ and some, rather small, set of apps installed as for this setup?
It evolved to that set as some helpful user-level utilities were not found at xfce-level.
Well, xdg-desktop-portal-gnome was removed in meantime. For sake of clarity pipewire, and those parts of pulseaudio are removed now too. As soon as problem is resolved the desire is to have pipewire as it seems to be newer state of engineering as pulseaudio.
This way the session freeze-up trigger has now less chance to kick in. Indeed, it occurs now less frequently, anyhow I believe to have got it once since removal of xdg-desktop-portal-gnome.
For further troubleshooting of user session total freeze-up I think to open new threat as there is no evidence the root cause is located in desktop environment as well as error occurrence last minute left in journal tracks of different system area. Follower topic is open.
It appears to me like you are looking towards building your own ISO with a package set more suited at your preference.
It is fairly easy to do - and the community have worked out a couple of documents and videos explaining the process. The following is a short introduction with links to more viewing
There is two text files in the systemroot which is lists of packages provided by the ISO experience.
roofs-pkgs.txt (common to all ISO)
desktopfs-pkgs.txt (environment specific)
The topic on [root tip] [How To] Do a manual Manjaro installation describes a method to parse those files and install them to a base system - but the method can be appiled for other use cases as well - such as reworking a running installation.
Thanks for input.
Actually it is a clone of virtual machine where user desire was to have Manjaro free of audio components. Hence as for clone source lot of audio packages were removed, if not all.
As far as it concerns this clone audio support is wanted, hence some set of audio packages were installed, first based on pulseaudio, at later point of time maybe some dirty mix pulse audio and pipewire; at current point of time clean switch to pipewire.
It can well be that packages juggling conducted in past was made in unfortunate and dirty way.
Similar may apply to desktop environment - however this side there is long-term desire to stay on XFCE, so if GNOME packages were installed these were rather user applications and their dependencies.
There exists no desire to build ISO image. Rather a tayloring of this or that Linux machine.
When microphone is connected to host machine microphone user session becomes totally and irrecoverably freezed-up. Majority of session freeze-up error occurrences happen when user conducts some really unsophisticated handling with one of sound applications, e.g. simple recorder, or pulse mixer. There are infrequent occurrences out of sound use-case, however still only while host machine shares mic with this Manjaro guest - as of this minute it is hard to describe circumstances of occurrences of latter type. Not a single freeze-up error was observed when host machine doesn’t share mic with Manjaro on guest machine.
Below journal log for period of time corresponding with latest error occurrence. Eventually cut of log is too radical, let me know please. This is tail of logs that one boot and first minutes of user session. I believe user attempted to open pulse mixer as it happened. At that point of time manjaro-pipewire was installed. As of this minute there is no desire to remove pipewire.
How bad for user session which is expected to continue its operation is what log excerpt below presents?
Jun 20 08:46:55 mach5 at-spi-bus-launcher[1247]: dbus-daemon[1247]: Activating service name='org.a11y.atspi.Registry' requested by ':1.0' (uid=969 pid=1236 comm="/usr/bin/lightdm-gtk-greeter")
Jun 20 08:46:55 mach5 dbus-daemon[846]: [system] Activating via systemd: service name='org.freedesktop.home1' unit='dbus-org.freedesktop.home1.service' requested by ':1.21' (uid=0 pid=1272 comm="lightdm --session-child 13 20")
Jun 20 08:46:55 mach5 dbus-daemon[846]: [system] Activation via systemd failed for unit 'dbus-org.freedesktop.home1.service': Unit dbus-org.freedesktop.home1.service not found.
Jun 20 08:46:55 mach5 lightdm[1272]: pam_systemd_home(lightdm:auth): systemd-homed is not available: Unit dbus-org.freedesktop.home1.service not found.
Jun 20 08:46:55 mach5 at-spi-bus-launcher[1247]: dbus-daemon[1247]: Successfully activated service 'org.a11y.atspi.Registry'
Jun 20 08:46:55 mach5 at-spi-bus-launcher[1274]: SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry
Jun 20 08:46:56 mach5 cups.cupsd[1280]: + /snap/cups/950/bin/lpstat -h /run/cups/cups.sock -r
Jun 20 08:46:56 mach5 cups.cupsd[1281]: + grep -qv not
Jun 20 08:46:56 mach5 cups.cupsd[892]: + sleep 1
Jun 20 08:46:57 mach5 cups.cupsd[1283]: + /snap/cups/950/bin/lpstat -h /run/cups/cups.sock -r
Jun 20 08:46:57 mach5 cups.cupsd[1284]: + grep -qv not
Jun 20 08:46:57 mach5 cups.cupsd[892]: + sleep 1
Jun 20 08:46:58 mach5 cups.cupsd[1286]: + /snap/cups/950/bin/lpstat -h /run/cups/cups.sock -r
Jun 20 08:46:58 mach5 cups.cupsd[1287]: + grep -qv not
Jun 20 08:46:58 mach5 cups.cupsd[892]: + sleep 1
Jun 20 08:46:59 mach5 cups.cupsd[1289]: + /snap/cups/950/bin/lpstat -h /run/cups/cups.sock -r
Jun 20 08:46:59 mach5 cups.cupsd[1290]: + grep -qv not
Jun 20 08:46:59 mach5 cups.cupsd[892]: + sleep 1
Jun 20 08:47:00 mach5 cups.cupsd[1293]: +
Jun 20 08:47:00 mach5 cups.cupsd[1292]: +
Jun 20 08:47:00 mach5 cups.cupsd[1293]: grep -qv not
Jun 20 08:47:00 mach5 cups.cupsd[1292]: /snap/cups/950/bin/lpstat -h /run/cups/cups.sock -r
Jun 20 08:47:00 mach5 cups.cupsd[892]: + sleep 1
Jun 20 08:47:01 mach5 cups.cupsd[1295]: +
Jun 20 08:47:01 mach5 cups.cupsd[1296]: + grep -qv not
Jun 20 08:47:01 mach5 cups.cupsd[1295]: /snap/cups/950/bin/lpstat -h /run/cups/cups.sock -r
Jun 20 08:47:01 mach5 cups.cupsd[892]: + sleep 1
Jun 20 08:47:02 mach5 cups.cupsd[1298]: +
Jun 20 08:47:02 mach5 cups.cupsd[1299]: + grep -qv not
Jun 20 08:47:02 mach5 cups.cupsd[1298]: /snap/cups/950/bin/lpstat -h /run/cups/cups.sock -r
Jun 20 08:47:02 mach5 cups.cupsd[892]: + sleep 1
Jun 20 08:47:02 mach5 systemd[1]: NetworkManager-dispatcher.service: Deactivated successfully.
Jun 20 08:47:03 mach5 cups.cupsd[892]: + [ 0 = 1 ]
Jun 20 08:47:03 mach5 cups.cupsd[892]: + wait 1097
Jun 20 08:47:03 mach5 kernel: 06:47:03.524183 timesync vgsvcTimeSyncWorker: Radical guest time change: 7 211 851 419 000ns (GuestNow=1 687 243 623 524 138 000 ns GuestLast=1 687 236 411 672 719 000 ns fSetTimeLastLoop=true )
Jun 20 08:47:22 mach5 systemd[1]: systemd-hostnamed.service: Deactivated successfully.
Jun 20 08:47:24 mach5 systemd[1]: systemd-timedated.service: Deactivated successfully.
Jun 20 08:48:26 mach5 lightdm[1272]: gkr-pam: unable to locate daemon control file
Jun 20 08:48:26 mach5 lightdm[1272]: gkr-pam: stashed password to try later in open session
Jun 20 08:48:26 mach5 systemd[1]: Stopping Session c1 of User lightdm...
Jun 20 08:48:27 mach5 lightdm[1221]: pam_unix(lightdm-greeter:session): session closed for user lightdm
Jun 20 08:48:27 mach5 systemd[1]: session-c1.scope: Deactivated successfully.
Jun 20 08:48:27 mach5 systemd[1]: Stopped Session c1 of User lightdm.
Jun 20 08:48:27 mach5 lightdm[1272]: pam_unix(lightdm:session): session opened for user kkkk(uid=1000) by kkkk(uid=0)
Jun 20 08:48:27 mach5 systemd-logind[850]: Removed session c1.
Jun 20 08:48:27 mach5 systemd[1]: Created slice User Slice of UID 1000.
Jun 20 08:48:27 mach5 systemd[1]: Starting User Runtime Directory /run/user/1000...
Jun 20 08:48:27 mach5 systemd-logind[850]: New session 2 of user kkkk.
Jun 20 08:48:27 mach5 systemd[1]: Finished User Runtime Directory /run/user/1000.
Jun 20 08:48:27 mach5 systemd[1]: Starting User Manager for UID 1000...
Jun 20 08:48:27 mach5 dbus-daemon[846]: [system] Activating via systemd: service name='org.freedesktop.home1' unit='dbus-org.freedesktop.home1.service' requested by ':1.26' (uid=0 pid=1310 comm="(systemd)")
Jun 20 08:48:27 mach5 dbus-daemon[846]: [system] Activation via systemd failed for unit 'dbus-org.freedesktop.home1.service': Unit dbus-org.freedesktop.home1.service not found.
Jun 20 08:48:27 mach5 (systemd)[1310]: pam_systemd_home(systemd-user:account): systemd-homed is not available: Unit dbus-org.freedesktop.home1.service not found.
Jun 20 08:48:27 mach5 (systemd)[1310]: pam_warn(systemd-user:setcred): function=[pam_sm_setcred] flags=0x8002 service=[systemd-user] terminal=[] user=[kkkk] ruser=[<unknown>] rhost=[<unknown>]
Jun 20 08:48:27 mach5 (systemd)[1310]: pam_unix(systemd-user:session): session opened for user kkkk(uid=1000) by kkkk(uid=0)
UPDATE: Few minutes later one more test with same failure and again timesync vgsvcTimeSyncWorker: Radical guest time change. I think I need to face this direction.
UPDATE 2: Virtualization host keeps two following features enabled for this Manjaro station: KVM paravirtualization interface and nested paging hardware virtualization level. Former one shows in VirtualBox manual common points with failure log - time control. Administrator/user of Manjaro hesitates to disable it as hypervisor vendor recommends to use it. Nested paging disabled was in meantime tested out - no improvement.
Among those settings of virtual machine found at hypervisor level it is para-virtualization which shows some relation (based on its description, VirtualBox manual) to error log found in journal. VirtualBox maker doesn’t recommend to disable para-virtualization, hence administrator hesitates to conduct turn-off.
Disablement of nested paging results on another side in no improvement - one test was done.
Then reply to the same thread instead of creating a splintered, duplicate thread. Please be clear and concise about what you need help with to begin with.
Manjaro from-scratch install using 22.1.3 ISO downloaded from Manjaro download web page. Then upgraded to 23.0.0 using Add/Remove Software app.
In the end bunch pretty sized of Explicitly installed apps is presented by Add/Remove Software app when neither admin nor user installed any app by themself. Why is it this way?
Admins/users expectation is all apps aren’t presented as Explicitly installed but as provisioning by system / Manjaro distribution makers.
Also two apps are listed in Orphans category - admins/users expectation is package management had cleaned up orphans as for presented installation history.
“excplicitly installed” is a package marker that works, it appears, a little bit differently than you are expecting. It doesnt mean ‘installed by you yourself’ it means ‘particularly installed on purpose’ … so any package named in the initial package lists are ‘explicitly’ installed. If, for example, a dependency is required or becomes required for one of those packages - then that dependency package will not be marked ‘explicit’.
If you were to remove the ‘explicitly installed’ marker from all of these pre-installed packages they may be considered orphans and easily accidentally removed by any number of functions because they would not be considered ‘explicitly installed’.
I absolutely don’t understand why the last discussions merge was conducted. The discussion open today regarding Explicitly installed apps on almost virgin 23.0.0 installation is not needed in this discussion nor supports it.
Quite interesting finding was made few minutes ago in this regard:
Sleek new Manjaro installation using 22.1.3 ISO then all pending updates were installed using Add/Remove Software. No other software composition of Manjaro installation in mind was conducted.
In the end both xdg-desktop-portal-gtk and xdg-desktop-portal-gnome are installed that machine.
I’m not sure why.
Usually people have one of the portal pulled in as a dependency of something like flatpack support - it follows that *gnome is installed as the first presented dependency in alphanumerical order … either the user does not pay attention and accepts the first suggestion, or through whatever invocation of pamac are not presented with any option at all. xdg-desktop-portal-gtk is even preinstalled on contemporary XFCE ISOs:
VirtualBox implements a sane set of defaults based on the host hardware. It is not recommend to alter those unless you have explicit knowledge of how the changes will affect the Virtual Machine.
I can - to a degree - understand if you are confused about explicit installed packages as the understanding may be obscured by language barrier.
The desktop portal implementation is used by sandboxing features such as snap and flatpak, to pickup certain system settings otherwise unavailable due to the sandboxing.
From your log as quoted above - it appears you have installed a snap package which communicates with the system to pick up printing capablities.
When a package has no direct relation or indirect relation to other packages it is considered orphan - as in - it has no parents.
The desktop portal packages is mutally exclusive so one speculates how you managed to have both xdg-desktop-portal-gtk and xdg-desktop-portal-gnome installed
Just verified - it appears both are included with the full ISO - while none is included with the minimal - it has also been corrected for the next ISO
So if you haven’t already done so - you can safely remove it from your XFCE installation