I recently bought a new laptop and installed Debian 12 + xfce environment. And I'd like to access my server (which has Debian 12) with Thunar, as I always did before. I can access my server via sftp with the command line, but when I try to access it with the address bar in thunar (e.g. sftp://user@server.com/) it enters in a password loop, asking for the password continuously. Note that it should not even ask for the password, because I usually keep a ssh connection with the server and I use openssh multiplexing. I have all gvfs packages installed (needless to say that I can enter with my old laptop having ubuntu 20.04). Any clue? Thanks in advance for any suggestion. Alex

I'm using Ubuntu 22.04 with Gnome and **I want to create a bash script which copy files from my phone to my laptop**. Seeing this answer , I installed gvfs-fuse , reboot and now the device is available at /usr/run/$(id -u)/gvfs. The problem is that I want to mount MTP devices without needing gvfs, since I want the script to be usable in any linux environment. I tried using jmtpfs but it yields errors related to GVFS: - mtp-detect gives the error:

Attempting to connect device(s)
libusb_claim_interface() reports device is busy, likely in use by GVFS or KDE MTP device handling alreadyLIBMTP PANIC: Unable to initialize device
Unable to open raw device 0
OK.

- Trying to mount with jmtpfs yields (understandably) the same result:

libusb_claim_interface() reports device is busy, likely in use by GVFS or KDE MTP device handling alreadyLIBMTP PANIC: Unable to initialize device
terminate called after throwing an instance of 'MtpErrorCantOpenDevice'
  what():  Can't open device
fish: Job 1, 'jmtpfs /media/santatra/mtp/' terminated by signal SIGABRT (Abort)

- but jmtpfs -l see the MTP device:

Device 0 (VID=18d1 and PID=4ee1) is a Google Inc Nexus/Pixel (MTP).
Available devices (busLocation, devNum, productId, vendorId, product, vendor):
1, 36, 0x4ee1, 0x18d1, Nexus/Pixel (MTP), Google Inc

So it seems that the desktop environment is mounting the device automatically. **QUESTION**: How can I mount MTP devices without needing GVfs in an environment with GNOME? And how can I tell Nautilus to release the MTP device? ## What I tried - The *Unmount* button on Nautilus does indeed unmount the device but gives the same error. - There's no option related to automatic mounting in gnome-control-center ## Environment - Ubuntu 22.04 - Nautilus 42.2 ## NOTE I don't think I should ask on Ask Ubuntu because - it can also occur in a KDE environment - I want the script to be usable in any linux environment

Like many other posts, gio mount for me results in volume doesn't implement mount. I've followed just about every SE post for instantiating a dbus session prior to mounting but end up with the same result. Straight from the [source](https://wiki.gnome.org/Projects/gvfs/doc#:~:text=Active%20GVfs%20mounts%20are%20only,to%20start%20any%20daemon%20explicitly.) , I'm told: > The only requirement is a running user d-bus session, typically > started with desktop session on login. When using GVfs in an isolated > shell (console, ssh login) make sure to start the d-bus session (the > dbus-launch command, e.g.

dbus-launch

). Active GVfs mounts > are only available to the specific session, making it work nicely in > concurrent environment like multi-seat systems or terminal servers. > > Everything is autostarted on-demand on first access through GIO. No > need to start any daemon explicitly. Some mounts may require > credentials, this is done using standard GMountOperation object. > > All daemons are listening to session bus changes and will exit once > the bus is torn down. Should this not be true, please file a bug. Which also produces the same result. Are there any other causes for this error?

When in nautilus or caja I click on the icon of an encrypted disk and enter my password, the underlying block device gets mapped to /dev/mapper/luks-$UUID and it gets mounted at/media/$USER/$DISK, no root password required. Is there a way to invoke this process from the command line, without GUI, including obviating sudo and having the mountpoint able to get unmounted again from GUI.

I usb-connect my ios device on my linux laptop. The linux system mounts a gvfs point to it (/run/user/1000/gvfs/afc:host=xxxxxxxxxx,port=3), I change directory there, I can rsync back and forth, but I wish to clone a git repository who lives in my laptop (/home/user/repo.git).

clone /home/user/repo.git

It will mist-behave:

Cloning into 'repo'...
error: chmod on /run/user/1000/gvfs/afc:host=xxxxxxxxx,port=3/com.lhy.iMarkdown/repo/.git/config.lock failed: Operation not supported
fatal: could not set 'core.filemode' to 'false'

And, sudo git clone /home/user/repo.git will give fatal: could not create work tree dir 'repo': Permission denied. --- This explanation by @torek (https://stackoverflow.com/users/1256452/torek) of git clone answering the same situation but for a samba mount point: https://stackoverflow.com/a/50242563 Will say the process fails on the third and last step when git clone uses invokes the OS service to rename the file. --- In this case, GVfs mounts a removable foreign operating system that has storage (https://askubuntu.com/a/1035122) . This is why sudoing does not work as GVfs makes the mount point usable only by the user who mounted it. --- It would be very comfortable to connect my ios device to my linux laptop and use git in it to repositories in my local device/network. Does anyone know how to make this smooth, please? It would be good to leave the configuration in /etc/fstab so that the whole interaction be as natural as any other storage point that I plug and use. Please it does not make sense to rsync things that are on a git repository back and forth and re-create git. People might have worked around such issues using git on restrictive environments/boxes/devices/storages.

What does gvfs do for me on my Kubuntu machine and why is /usr/lib/gvfs/gvfs-gdu-volume-monitor eating so much CPU time? BTW: I read https://en.wikipedia.org/wiki/GVFS and still don't know what's in it for me, especially on KDE / Kubuntu. lsof shows me that thunderbird, firefox and pidgin have gvfs libraries open, but for what functionality?

I am trying to run a bash script on certain folder inside an Android phone, mounted in my Fedora Linux box. I can see that the phone is there with adb devices (no, I do not want to use ADB for this!), I can see the phone folders using dolphin, I can even see that the phone is there with mount, showing: kio-fuse on /run/user/500/kio-fuse-JhdSvC type fuse.kio-fuse (rw,nosuid,nodev,relatime,user_id=500,group_id=500) Yet, the folders /run/user/500/kio-fuse-JhdSvC, /run/user/500/gvfs, /run/user/500/gvfsd and ~/.gvfs are all empty, even when I see the folders in Dolphin. I also tried to mount the phone as I used to, with the command simple-mtpfs -o allow_root -o enable-move /mnt/phone, but I get: LIBMTP PANIC: Trying to dump the error stack of a NULL device So: how can I know in which folder in my filesystem the thing is mounted? Why is the folder /run/user/500/kio-fuse-JhdSvC empty? **EDIT**: Following this advice I am able to prevent KDE from automouting the Android Device as mtp. Yet, the popup in the phone to "Allow USB..." disappears as soon as it appears, so I cannot click on it.

So leaving details aside last night i stumbled across "administrator root" in a live instance of Zorin. First off, is Administrator Root even a thing? If so, where does it fall into the hierarchy? As i understood it there were three users in linux - root, admin, and user. Picture is attached. Thank you.

I can connect to network drives with the following command: ( echo myusername; echo MYDOMAIN; echo password123 ) | gio mount smb://127.0.0.1/Sharename And I can get some file attributes of the mount point: gio info smb://127.0.0.1/Sharename Anzeigename: Sharenameauf 127.0.0.1 Name bearbeiten: / Name: / Typ: directory Größe: 0 Adresse: smb://127.0.0.1/Sharename/ Attribute: standard::type: 2 standard::name: / standard::display-name: Sharename auf 127.0.0.1 standard::edit-name: / standard::icon: folder-remote, folder-remote-symbolic standard::content-type: inode/directory standard::size: 0 standard::allocated-size: 0 standard::symbolic-icon: folder-remote-symbolic, folder-remote etag::value: 0 id::filesystem: smb-share:server=127.0.0.1,share=Sharename access::can-trash: FALSE time::modified: 0 time::modified-usec: 0 time::access: 0 time::access-usec: 0 time::changed: 0 time::changed-usec: 0 unix::device: 2648392497 unix::inode: 6160419607250089214 But how can I fetch the name of the user that connected to the share ("myusername" in the case above)?

Looking for an alternative to gksudo in Debian, I stumbled upon [this article](https://itsfoss.com/gksu-replacement-ubuntu/) , which suggests to use the admin:// prefix in order to access files as root. E.g. instead of gksu gedit /etc/default/apport one is supposed to run gedit admin:///etc/default/apport I don't use gedit, but running leafpad admin:///etc/fstab on my Debian Buster simply opens an empty file. What am I missing? mtp:// is working on that machine, so I suppose GVFS support is present.

I am working on a remote server that runs CentOS 7.4.1708. On this server, I managed to mount a fileserver, which is in the same network, with the following commands that I found somewhere on the internet:

export $(dbus-launch)
gvfs-mount smb://xxx.xxx.xxx.xxx/abc/

Now I was able to access the files at

/run/user/100*/gvfs/smb-share\:server\=xxx.xxx.xxx.xxx/\,share\=abc/

Once I finished moving files etc., I usually unmounted the server:

gvfs-mount -u smb://xxx.xxx.xxx.xxx/abc/

Now the problem is, that **I forgot unmounting and just closed the ssh connection** to the CentOS server. Once I logged in again, I could still access the fileserver at /run/user/100* and that directory still showed me as the owner. However, gvfs-mount -l does not show the fileserver anymore and trying to unmount it results in the following error message

Error finding enclosing mount: Containing mount does not exist

**Edit:** Remounting of the fileserver does not work, the gvfs-mount job does not finish within 10 minutes and seems to do nothing. **How can I unmount the fileserver now?** (I have no sudo rights)

I am running Linux Mint 17.1 64-bit (based on Ubuntu 14.04). Ever since upgrading from Linux Mint 14/Ubuntu 12.10, the Python script I use to sync music to my Walkman has stopped working. Previously, when I mounted my Walkman, it would automatically show up as the path /run/user/1000/gvfs/WALKMAN/Storage Media and would work like any other file system: I could copy tracks to it, delete tracks from it, etc, all through Python. However, I can't remember if I had to make any changes to get this to happen. Since upgrading to Linux Mint 17 (and now 17.1), when I mount the Walkman, it shows up as the path /run/user/1000/gvfs/mtp:host=%5Busb%3A002%2C007%5D/Storage Media. Furthermore, when I try to run the same file operations, they now fail. I have discovered that this happens not just through Python, but on the command line as well. For example: david@MILTON:~$ cp '/data/Music/10SecsWhiteNoise.mp3' '/run/user/1000/gvfs/mtp:host=%5Busb%3A002%2C006%5D/Storage Media/MUSIC' cp: cannot create regular file ‘/run/user/1000/gvfs/mtp:host=%5Busb%3A002%2C006%5D/Storage Media/MUSIC/10SecsWhiteNoise.mp3’: Operation not supported I have done some research on this problem, but the most common explanation seems to be that it was formerly solved by this PPA: https://launchpad.net/~langdalepl/+archive/ubuntu/gvfs-mtp But now, Ubuntu versions since 13.10 contain all these changes so it should no longer be necessary. So why am I still having these errors? I am still able to do file operations on my Walkman through a graphical file manager (Caja, on Linux Mint), just not via the command line.

Sorry for the question that is too general, but I would like to switch to using the terminal instead of GUI programs as much as possible and I cannot formulate my question more precisely. In the course of my work, I need to connect to multiple FTP directories and Samba shared drives and also for MTP (smartphone) drives at home. I want to join them in the same way so I can handle them as uniformly as possible. Currently, GIO Virtual Filesystem / GVFS seems to be the best solution because I can log in to the drives in a similar way (gio mount) and access them via a similar path (/run/user/1000/gvfs/...). The problem is that this solution seems to be optimized for GUI because e.g. I can't copy to MTP (smartphone) drives, only from the GUI, but the GUI also uses GVFS. Is there a proven way to connect to different (network) drives from a terminal as easily as from a GUI?

I use Krusader to access my Android phone and copy files to it. The path is this /run/user/1000/gvfs/mtp:host=/ Firstly, it won't display the content of that path, even though it's been mounted, unless I first access via Thunar. Only after that Krusader will display the content of a phone and its SD card. Secondly,when I'm attempting to copy a file, it'll throw an error "couldn't write into ... /downloads/123.txt" How to fix all that?

This is what I believe to be the root of a much larger problem with my install of GVFS. I'm running XFCE4 on Linux Mint 20.04, and from *day 1* have not been able to use GVFS functions in any file manager. I can't use trash, I can't access SFTP servers or anything that requires GVFS. Trying to open the trash on Thunar displays > **Failed to open 'trash:///'** Operation not supported. Similar messages appear on other file managers, usually with 'Operation not supported.' When going into preferences, Thunar displays > It looks like **gvfs** is not available Upon actually digging in the Thunar source code, I found that what it is actually doing here is testing the get_supported_uri_schemes function for common things like trash and sftp, and saying 'gvfs is not available' if those don't show as supported. I tested this myself with the below code snippet (taken from the Thunar source), and the only supported scheme returned by the function is file. I can confirm this by typing paths into the Thunar top-bar; it complains when anything except file:/// is entered. const gchar * const *supported_schemes; gboolean supported = FALSE; guint n; GVfs *gvfs; gvfs = g_vfs_get_default (); supported_schemes = g_vfs_get_supported_uri_schemes (gvfs); if (supported_schemes == NULL) return FALSE; for (n = 0; !supported && supported_schemes[n] != NULL; ++n) if (g_strcmp0 (supported_schemes[n], "trash") == 0) supported = TRUE; return supported; Upon even further digging into old forum posts , it seems that on a working machine, the gvfs daemon is supposed to be started automatically by Nautilus, Thunar or any other file manager that requests it (not sure how that works). This doesn't happen on my machine, for a reason I've yet to determine. I am able to start the service manually using systemctl --user start gvfs-daemon, but this process running doesn't change the output of get_supported_uri_schemes. **Things I've Tried (that didn't work)** 1. Reinstalling glib, gvfs (and every listed dependency for gvfs package) 2. Installing gvfs-backends, gvfs-common etc. 3. Adding exec dbus-launch --sh-syntax --exit-with-session xfce4-session to my .xinitrc (as per this old post ) 4. Manually starting gvfs-daemon using systemctl 5. Running the entire session as root (startx from root terminal), the same problem there Can anyone point me in the vague direction of what might be wrong with gvfs here? Why doesn't it register as supporting anything other than file:///? Why doesn't the daemon start automatically with the file managers? I appreciate any help anyone can render here, please comment if more information is required.

I use Gnome 3.36.3 (Ubuntu 20.04.1) with GnuPG 2.2.19 as SSH agent. I tell OpenSSH where to find the GnuPG SSH agent socket in my

~/.profile

file: export SSH_AUTH_SOCK=$(gpgconf --list-dirs agent-ssh-socket) This works fine in terminal windows. Whenever I connect to a remote SSH server, I get prompted for my PIN. But when I try to connect to a remote SFTP server with Nautilus, I get errors, either an message like "not authorized!" or I am asked for username and password (which of course won't work as I use keys only). As a manual workaround I found out, that after a

gvfsd

and a complaining beep from Nautilus, SFTP connections will work as expected. So apparently **gvfs** is not aware of my GnuPG SSH agent after login. but will be later, when killed and restarted. What do I have to do, to make **gvfs** aware of my GnuPG SSH agent, wwithout killing manually first?

I can list the mimetypes potentially handled by a .desktop file doing something like this: $ grep 'MimeType' /usr/share/applications/eog.desktop | tr ';' '\n' MimeType=image/bmp image/gif image/jpeg image/jpg image/pjpeg image/png image/tiff ... And I can show what the default application is for a given mimetype with gvfs-mime: $ gvfs-mime --query image/tiff Default application for 'image/tiff': eog.desktop Registered applications: eog.desktop gimp.desktop ... Recommended applications: eog.desktop gimp.desktop ... or with gio: $ gio mime image/tiff Default application for “image/tiff”: eog.desktop Registered applications: eog.desktop gimp.desktop ... Recommended applications: eog.desktop gimp.desktop ... But how do I list the mimetypes that eog.desktop actually handles?

My desktop is using OpenSuse 15.1 with latest updates, systemd 234-lp151.26.16.1, gvfs 1.34.2.1-lp151.6.3.1. Since a while (I cannot remember exactly when it started) a system shutdown takes at least 2 minutes to complete. When I hit ESC at the splash screen, I see systemd is waiting for a User Manager job to finish and, after some digging, I figured out that it is this process: /usr/lib/gvfs/gvfsd-fuse /run/user/1000/gvfs -f -o big_writes I did some research and discovered that this is automatically started by programs like Firefox and Thunderbird (and others). I also verified that: - If I boot and don't use such programs, the shutdown is fast; - If I logout and use the console to kill those user processes, the shutdown is fast; - Other user accounts are affected too, only if they use the mentioned programs. I configured systemd with: KillUserProcesses=yes, but it does not help as it seems those processes still take 2 minutes to be terminate after the logout. Furthermore, on the same machine I can also boot Ubuntu-mate (18.04 LTS) from a separate SSD and it shows the same symptoms. How do I figure out what's causing the 2 minutes delay ? Edit: perhaps it's worth mentioning that once I logout, I cannot log back in again because I get a dialog "Could not sync with Dbus". I need to do a ctrl-Backspace to reset X first.

I would like to be able to click on SSH links that look like ssh://user@servername and have a gnome-terminal open that runs SSH inside of it. After reading up on the XDG stuff I did the following: I created a shell script in $HOME/bin/ssh-terminal with the following: #!/bin/sh dest=${1#ssh://} gnome-terminal -e "ssh $dest" & This works from the command line. I created $HOME/.local/share/applications/ssh-terminal.desktop with: [Desktop Entry] Type=Application Name=SSH Terminal Exec=ssh-terminal %u Icon=utilities-terminal StartupNotify=false MimeType=x-scheme-handler/ssh; I ran update-desktop-database ~/.local/share/applications/ Now when I try to run xdg-open ssh://user@hostname I get the following error: $ xdg-open ssh://buildbot@10.16.50.74 gvfs-open: sftp://buildbot@10.16.50.74/: error opening location: The specified location is not mounted I don't *want* to open it as a sftp mount and I cannot seem to find where that is set! Where do I go from here?

I am new to Linux so excuse my noob. I haven't been able to find an answer for this anywhere nor have I even found anyone with a similar problem and no one on discord has an answer either.. I am running Arch with DWM and Kittie. I have installed the py and alsa packages. When I run Blender from dmenu nothing happens. Running it in terminal launches blender for a second with a blank or bugged out screen before crashing and returning: /run/user/1000/gvfs/ non-existent directory Floating point exception (core dumped) I tried running the command "sudo blender": ALSA lib pcm_dmix.c:1089:(snd_pcm_dmix_open) unable to open slave I made the gvfs directory from the command line, reinstalled Asla and rebooted and now I just get the Floating point exception both ways. Any ideas? Note: every time I reboot for anything the gvfs file disappears.. this is a genome file dir if I'm not mistaken so I'm even more confused now. Update : I attempted to remove, install Steam and install Blender through stam and it still fails with the folowing error Exiting app 365670 No cached sticky mapping in ActivateActionSet.