Frank Nocke's questions

Inspired from this topic I dare ask this probably impossible question:

I have 4 different users on my Ubuntu-MATE 21.04 machine. Sometimes it would be nice to see user2 on the display of user1, either in a Window (some sort of exported X-Window, rlogin,...) or even better, one of those 4 Workspace Panels:

So like Panel 1-3 for user 1, Panel 4 for user 2. (I am aware that I can do sudo - …, rlogin, ssh on the command line to run certain commands as another user, but now+then I would truly need that other use with the actual GUI...)

If there is a solution, perhaps it is in the world of export DISPLAY=:0.0 NO_AT_BRIDGE=1 or in this answer. Perhaps exporting is easy and I just need a rather dumb X-Display client? As you can tell, I have very little clue around Displays, XServer and Exports.

I have used a keyboard short calling shutter -s and making an rectengular area screenshot under Ubuntu-MATE 16.04 for many years. No problems ever. Of course, shutter was already running... the redundant call would simply do it's job of captuing (and or tell the original instance, either way, no problem, proper capture, proper display in shutter window)

Now, under Ubuntu-MATE 21.04 and with shutter 0.94.3 this is no longer working. Running shutter -s get's me:

ln: failed to create symbolic link '/home/frank/snap/shutter/27/.config/gtk-`enter code here`2.0/gtkfilechooser.ini': File exists
Gtk-Message **: Failed to load module "appmenu-gtk-module" at /snap/shutter/27/usr/lib/x86_64-linux-gnu/perl5/5.22/Gtk2.pm line 126.
WARNING: gnome-web-photo is missing --> screenshots of websites will be disabled!


(shutter:74252): Unique-DBus-WARNING **: Error while sending message: Message recipient disconnected from message bus without replying

INFO: There is already another instance of Shutter running!

• „screenshots of websites will be disabled“ has been around for years and should not be related
• not so sure what this symbolic link-warning is about...
• maybe the very last „INFO“ is a problem now, maybe it isn't?

What does work 90% of the time (as a script in keyboard shortcuts) is this workaround, but hey...

sh -c "killall shutter ; sleep .2; shutter --delay=2 -s &"

... ist there a better way to do this? Or a known bug on shutter's end?

(there were some suggestions out there to use shutter -s -e, but it is not helping)

I want every USB media (ext, ntfs, FAT…) inserted to be available with full read-writable to every user. Preferably under mnt and as group common.

( Ubuntu-MATE 21.04, though my issues actually date back even to Ubuntu-MATE 16.04...)

1. I boot into my machine as user frank. When I mount an USB drive (FAT) it ends up under /media/frank/4C8C-E4BC – just fine, as it should. When I switch to another user (keeping the first session open, not logging off, i.e. dm-tool switch-to-user lisa) I get

Unable to mount 16 GB Volume Device /dev/sda1 is already mounted at /media/lisa/4C9C-xxxx

Making sense so far. But not what I want.

Sure, for individual drives I could change the mount location via gnome-disks or manually add a line to etc/fstab (which essentialy is what gnome-disks is doing). However I do not want to do this for each and every USB stick and SD card I posess...

Is there a better way for an all-user automount?

[ I understand/guess there's something called autofs handles the automatic mounting of USB sticks? But as far as I can tell, that's no part of my Ubuntu-MATE install... ]

And then there seems to be “udev-Rules”, is that, what I want? Apparently it's even an “inevitable” part of Ubuntu, as it is part of systemd (?)? There are rules on my machine under /etc/udev/rules.d although seemingly rather specific ones about brave and skype and signal...)

There's promising stuff MODE and GROUP and UDISKS_FILESYSTEM_SHARED but (provided this is the right approach) I can't set the ultimate puzzle pieces together. It should truly only apply to removable drives like USB sticks and memory cards, otherwise I get into deep trouble... this source mentions SUBSYSTEMS=="usb", that would be a good filter, if true...)

Is udev rulez the way?

If yes, can somebody set the puzzle pieces together? I am almost suspecting all I look for is a one-liner file in /etc/udev/rules.d.

addendum:

Not my main problem, but if it went away for the same reasons, I would be happy: Inserting a Windows Boot USB Stick (slightly more commplex, comes with an UEFI Boot FAT and a Windows Install NTSF partition) under the second user, I instantly get:

(and no, intense googling didn't get me anywhere) …switching back to first user frank (stick still inserted) I am welcomed by these authorization requests:

addendum:

Not a fix, but if those two users are essentially you, and the other one is not truly active (just switched-away from instead of logged-out) this is the quickest way before a trouble-free USB mount:

who -u
sudo kill <that other user's pid>

I am using Libreoffice 7.1.4.2 under Ubuntu-MATE 21.04. According to the LO's about dialog: User Interface: UI render: default; VCL: gtk3

I only use the text processing „Writer“, nothing else.

Problem 1: my styles panel (on the right) on the right flickers all the time

...blinks, maybe 50-100ms to white and back... (but certainly not in a blink tag fashion, more of a “nervous redraw”)

• basically irregulary about once per second (1 Hz)
• more often with every keypresses happening
• not at all, if focus is on another application's windows
• only in the styles panel

So apparently triggered by some refresh mechanism.

Problem 2

I assume it's related, thus in one post. LibreOffice freezes my entire machine:

• mouse still movable, all screens showing (3 in total, including 2× 4K)
• that's all – no clicks, no hovers, no interaction
• no keyboard commands (i.e. for Run or Terminal or alt-Tab) working
• not even ctrl-alt F3, F6, F7 (for non-graphical login)
• only hard reset + reboot helps

Any ideas?

What log file to look into? How to put LibreOffice into a different rendering mode?

I am on an unfortunate text adventure getting to run VirtualBox 6.1.22 under Ubuntu MATE 20.04 LTS. My key problem is, that I had to upgrade my (shiny new) machine to Kernel 12.4 (shiny new graphics card and other hardware requiring a higher kernel...)

frank@p14:~$ uname -r
    5.12.4-051204-generic

I followed the official Virtual Box instructions (adding to sources.list, adding key) and manage to install version 6.1.22, which explicitly supports Kernel 5.12..

( SecureBoot in BIOS is OFF. Just saying. )

The warning I get during install is:

This system is currently not set up to build kernel modules. Please install the Linux kernel "header" files matching the current kernel for adding new hardware support to the system.

The distribution packages containing the headers are probably:

linux-headers-generic linux-headers-5.12.4-051204-generic

...fair enough, and also upon starting $> virtualbox I get...

WARNING: The vboxdrv kernel module is not loaded. Either there is no module available for the current kernel (5.12.4-051204-generic) or it failed to load. Please recompile the kernel module and install it by

sudo /sbin/vboxconfig

You will not be able to start VMs until this problem is fixed.

...and indeed inside the Virtualbox GUI I can do little more than opening the 'about' box...

Those headers appear available here, very first link even:

**However they won't install without a libc6 2.33 **

Fiddling the hard way with libc6 versions sounds not like a good idea, or is it?

Are those “headers” available as a binary elsewhere? (my poor memories of C/C++ coding would tell me, that header files itself don't need compilation, but well in this context I have no clue what they are anyway...)

If this is the valid method to find out, I am on libc verison 2.31, so just two microversions below.

> ldd --version
>     ldd (Ubuntu GLIBC 2.31-0ubuntu9.2) 2.31

So behaps I may dare upgrade? Or is this dangerous? Or fiddle with the header package (how?) to accept 2.31? (wildly guessing there is no vital feature hidden in that minor upgrade...)

Addendum

$> sudo apt install linux-headers-generic
    linux-headers-generic is already the newest version (5.4.0.77.80).

...and /sbin/vboxconfig keeps complaining as above (since 5.4.0 is below 5.4.12).

Trying to install the required explicit version explicitly fails, but with some additional info:

sudo apt install linux-headers-5.12.4-051204-generic
    ...
    Package linux-headers-5.12.4-051204-generic is not available, 
    but is referred to by another package.
    This may mean that the package is missing, has been
    obsoleted, or is only available from another source

Hmm... should I maybe add another repo to /etc/apt/sources.list ?

Any other suggestions?

I have a shiny new Lenovo P14s 2nd gen ( installed with dual boot, Win10 and… ) Ubuntu MATE 20.04.2 LTS. And I can't connect by Wi-Fi. For some graphics-related driver issues (a little too shiny new after all for Ubuntu 20.04 it seemed) I did upgrade the kernel to 5.12.4 generic, but this is probably not relevanthere.

I can't see any network connections. Neither in the top panel (where a respective pulldown would sit in the Notification area with the well-known wifi waves-symbol) nor in the control-center (like I do under my old 16.04 machine, where there was a „Network Connections“ with all the nearby Wifi's to choose from) from. Here, it's missing:

(the usual rest is there, but nothing on “Network Connections”...)

I first suspected that also my Wifi adapter might be 'too' new and needs drivers… However I a) installed some drivers follwoing Lenovo documents and b) also plugged in a truly simple (3-4 year-old) USB LogiLink-Wifi adapter – no luck, no wifi connections offered anywhere.

inxi -N says, it's all there!

$> inxi -N
Network:    Device-1: Intel Ethernet I219-V driver: e1000e
            Device-2: Intel driver: iwlwifi
            Device-3: Ralink RT5370 Wireless Adapter type: USB driver: rt2800usb
            Device-4: Lenovo USB3.1 Hub type: USB driver: r8152

…and so does iwconfig. The wlx... one must be the external one (simple plug-in-plug-out test). I am not a driver/hardware expert, but looks to me like both devices are recognized and even operational (even with Bit Rate and Quality measuring under way...)

lo        no wireless extensions.
enp0s31f6  no wireless extensions.
enx083a885b1fdf  no wireless extensions.
wlp9s0    IEEE 802.11  ESSID:"WN-62F2C0"
      Mode:Managed  Frequency:5.54 GHz  Access Point: 7C:<redacted>:C5
      Bit Rate=866.7 Mb/s   Tx-Power=22 dBm
      Retry short limit:7   RTS thr:off   Fragment thr:off
      Power Management:on
      Link Quality=69/70  Signal level=-41 dBm
      Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
      Tx excessive retries:0  Invalid misc:63   Missed beacon:0
docker0   no wireless extensions.
wlx7cdd9017b62f  IEEE 802.11  ESSID:off/any
        Mode:Managed  Access Point: Not-Associated   Tx-Power=20 dBm
        Retry short  long limit:2   RTS thr:off   Fragment thr:off
        Power Management:off

$> dpkg -l | grep -i "network.*manager"

ii libnm0:amd64                          1.22.10-1ubuntu2.2 amd64 GObject-based client library for NetworkManager
ii libproxy1-plugin-networkmanager:amd64 0.4.15-10ubuntu1.2 amd64 automatic proxy configuration management library (Network Manager plugin)
ii network-manager                       1.22.10-1ubuntu2.2 amd64 network management framework (daemon and userspace tools)
ii network-manager-gnome                 1.8.24-1ubuntu3    amd64 network management framework (GNOME frontend)
ii network-manager-openvpn               1.8.12-1           amd64 network management framework (OpenVPN plugin core)
ii network-manager-openvpn-gnome         1.8.12-1           amd64 network management framework (OpenVPN plugin GNOME GUI)
ii network-manager-pptp                  1.2.8-2            amd64 network management framework (PPTP plugin core)
ii network-manager-pptp-gnome            1.2.8-2            amd64 network management framework (PPTP plugin GNOME GUI)

I am really clueless... thank you for your help!

I think I have basically understood how those systemd drop-in files work, like /etc/httpd/conf.d/ or /etc/sudoers.d/

My question: is there some helper command to show the current aggregate result?

i.e. seeing how everything plays together in the .d Folder and verifying i.e. that fooProperty is finally true not false…

You are looking at the boot stick for Ubuntu Mate 20.04. Plus a folder frankscripts containing a row of bash-scripts (to create multiple usersr, common folders, a bunch of chrome profiles, etc)

I wish to run those scripts as root after all else is done. (occasional user interaction like adding a password or two micht become necessary, but that's about it.

How can I turn this into an unattend install?

Booting from the stick, having everything done... (if possible, provide a script for my partitioning strategy, but just on a side note).

I did read the word kickstart in some places.

But I still have no clue what to do and where to start... how to enter or hook into the installer, or extend it (in whichever ways Ubuntu-Mate does or allows this).

I know that one can do quite much programatically (shell commands) to the attached monitor hardware in Linux:

• turn it on/off (truly, hardware-wise) with xset dpms force off
• toggle the brightness (again, truly hardware)
• toggling the backlight
• change gamma

So here comes the question:

Can I somehow (programatically) reach the White Balance/Color Temperature Settings of my monitor?

I am asking specifically for Samsung PM55F-BC and PM43F-BC. (Manuals confirm, there are at least options for manually doing it on the device...) We have plenty of kiosk devices and I'd rather avoid visiting each of them...

I am running Ubuntu MATE 20.04 and read through much theory around xkb and I managed for example to change altGr-, from · to (a bigger, bolder) •:

$> testing keys... öäü…·
$> cp de de_frank  (yes, coming from a German keyboard)

    OLD   key <AB08>  { [ comma,  semicolon, periodcentered,  multiply  ] };
    NEW   key <AB08>  { [ comma,  semicolon,          U2022,  multiply  ] };

$> setxkbmap de_frank
$> testing key... öäü…• good :)

Problem / Question:

I want to replace altGr-shift-9 from ± to ? (U1F600)

Does work on trivial keys like shift-3, but I don't know where I can find the right place in the file. Line :115 (in my case, Ubuntu MATE 20.04) has NO effect regarding the 9 key with any modifieres whatsoever:

< key <AE09> { [ 9, parenright, bracketright, NoSymbol, fiveeighths, plusminus, bracketleft, NoSymbol ] };

> key <AE09> { [ 9, parenright, bracketright, NoSymbol, U1F600, U1F600, U1F600, NoSymbol ] };

Putting UF1600 on, say, shift-3 (usually the dollar sign) is not a problem. So the Problem is not with the use of a rather exotic unicode.

I am guessing, the settings for altGr-shift-9 come from elsewhere, the „partial“ in which my change sits gets never included, or latter overriden, I do not exactly see through the relations and priority of all this...

possibly useful:

$>  setxkbmap -print -verbose 10

Trying to load rules file /usr/share/X11/xkb/rules/evdev...
Success.
Applied rules from evdev:
rules:      evdev
model:      pc105
layout:     de_frank
Trying to build keymap using the following components:
keycodes:   evdev+aliases(qwerty)
types:      complete
compat:     complete
symbols:    pc+de_frank+inet(evdev)
geometry:   pc(pc105)
xkb_keymap {
    xkb_keycodes  { include "evdev+aliases(qwerty)" };
    xkb_types     { include "complete"  };
    xkb_compat    { include "complete"  };
    xkb_symbols   { include "pc+de_frank+inet(evdev)"   };
    xkb_geometry  { include "pc(pc105)" };
};

I want to adjust a grub theme to have a truly large font (about double from what is normal). I usually work with the install.sh that comes with it and with grub-customizer.

The easy part, consensus everywhere: Take a (vector, i.e. ttf) font, and convert it to the (pixel, .pf2) font, placed in /boot/grub/fonts/:

sudo grub-mkfont --output=/boot/grub/fonts/DejaVuSansMono36.pf2 \
    --size=36 /usr/share/fonts/truetype/dejavu/DejaVuSansMono.ttf

The tough or strange part: Specifying this font in the central theme's settings file is not working. Somehow a space separated formed is required?!?

title-font: "Sans Regular 12"
or
title-font: "Gnu Unifont Mono Regular 16"
or
terminal-font: "Lato 12"

Questions:

• How do I get from above literal filename DejaVuSansMono36.pf2 to the "pretty name"? (guessing the right spots for spaces?)
• And how does the referencing work?
• And, should I put the .pf2 in the final place, or place it in theme folder? (looking at github, the latter seems right...)

The grub theme spec does not says anything about this <longname><space><size> format. Given that .pf2 are hardcoded pixel-fonts (like in the 90's :-) I wonder what value a size indication is anyway...

small addition

The pretty, “spacy” name can be found out by virtue of hex editor: Still a bit of a wonder, how it gets referenced. Does grub (or grub customizer) parse all existing *.pf2 files for the prettier name contained ?!? (or is there a lookup-table somewhere I… overlooked?)

I am using two different Lenovo laptops and even two different Ubuntu-MATE versions 16.04 and 18.04, yet ever since I started with Ubuntu, I suffer refresh / repaint problems (on a multitude of monitors and docking stations, so let's not blame those...)

If I switch to the next workspace (virtual screens) and back, these artifacts are gone. If I move a window slowly over it, it's also gone. Basically most things triggering a repaint makes it gone. Hastily shaking windows around tends to create them.

I know little about Linux & hardware, but let me share what I know. I appear to have intel graphics:

$> sudo lspci|grep VGA
    00:02.0 VGA compatible controller: Intel Corporation 3rd Gen Core processor Graphics Controller (rev 09)

And DRI2 (what that was again?)

$> xdpyinfo | grep DRI
    DRI2

My xorg config (if I am not talking gibberish) does not seem to contain a 20-intel.conf...

$> cat /usr/share/X11/xorg.conf.d/
10-amdgpu.conf            10-quirks.conf            11-evdev-trackpoint.conf  50-vmmouse.conf           51-synaptics-quirks.conf
10-evdev.conf             11-evdev-quirks.conf      50-synaptics.conf         50-wacom.conf

Not sure, what to make of that 10-amdgpu.conf file (Didn't I have intel graphics?)

cat /usr/share/X11/xorg.conf.d/10-amdgpu.conf
Section "OutputClass"
        Identifier "AMDgpu"
        MatchDriver "amdgpu"
        Driver "amdgpu"

Is this usr/share-xorg.conf.d just an overlay of settings, and the basic, core conf is elsewhere? Should I possible creatre a 20-intel.conf as suggest here.

Or is the problem somewhere else entirely? I am totally cluessless on these matters.

I have a weired sympton with a fresh Windos 10 guest OS under VirtualBox 5.1.28 under Ubutunu-MATE 16.04:

I can't move the window at all, when [x] Auto Capture Keyboard is on.

Neither dragging the outer windows borders (resize), nor the titlebar (indeed just move the virtuabl box. So this is not about auto-resize on or off...). If I deactivate the option, things work fine again. (I do not even understand the relation). – sounds like a virtual box bug? Can't reproduce the opposite, but I have been working with virtual box (and Win7 guests) for long, never encountered this until 2 weeks ago.

But then, common shortcuts I have defined in host and guest, don't work on the guest but rather the host (i.e. WIN-R to get a run programm...). So, of course capturing is desirable.

I am running Skype 8.20 under Ubuntu-MATE 16.04. It automatically starts after boot, but unlike defined in the settings: not minified / in-the-background, but as a regular window:

My question would be: Is there a way to ensure it starts minified? (Perhaps by finding the “startup folder“ of sorts under Ubuntu and adding a minify flag there? Wherever that is, perhaps the switches there are only compatible for KDE, Gnome, whatever… )

I am running Ubuntu MATE 16.04 and copying images to and from my internal SD card reader basically works fine… on my main account.

If another family member logs in (my login still active in the background, not logged out) and then inserts an SD card, he can't access it:

So, is this some kind of rights issue...? Logging back in as the main user, I also see some rights issues:

So, whatever it is, how can I loosen up access rights on the SD card reader, so it becomes accessible to everyone (including non-admin users)?

Perhaps some ‘relaxed’ etc/fstab entry? (no related entry there yet)

Update I:

Switching back & forth between 2 SD cards (from a single user) and looking at /dev/disk/by-id:

?> ~ $ ll /dev/disk/by-id | grep mmc
...  mmc-SDU1_0x02400f04 -> ../../mmcblk0
...  mmc-SDU1_0x02400f04-part1 -> ../../mmcblk0p1
?> ~ $ ll /dev/disk/by-id | grep mmc
...  mmc-SE32G_0x0072e5ab -> ../../mmcblk0
...  mmc-SE32G_0x0072e5ab-part1 -> ../../mmcblk0p1
?> ~ $ ll /dev/disk/by-id | grep mmc
...  mmc-SDU1_0x02400f04 -> ../../mmcblk0
...  mmc-SDU1_0x02400f04-part1 -> ../../mmcbl

?> ~ $ ll /dev/disk/by-id | grep mmc
...  mmc-SDU1_0x02400f04 -> ../../mmcblk0
...  mmc-SDU1_0x02400f04-part1 -> ../../mmcblk0p1
?> ~ $ ll /dev/disk/by-id | grep mmc
...  mmc-SE32G_0x0072e5ab -> ../../mmcblk0
...  mmc-SE32G_0x0072e5ab-part1 -> ../../mmcblk0p1
?> ~ $ ll /dev/disk/by-id | grep mmc
...  mmc-SDU1_0x02400f04 -> ../../mmcblk0
...  mmc-SDU1_0x02400f04-part1 -> ../../mmcblk0p1

../../mmcblk0[p1] ⇒ /dev/mmcblk0[p1] seems to be the only constant here. Perhaps I can get a solution with this?

I added this to /etc/fstab (according to man page, the first param may also be a device path, if I understand correctly):

/dev/mmcblk0p1  /media/ExtSD01/  auto   auto,user,rw 0 0

This worked somewhat nicely, now that I can mount /dev/mmcblk0p1/ and unmount to the user-neutral path, not automatically but at least from command line (disadvantage 1). However my system hang an reboot – unless I have an SD card inserted before boot (disadvantage 2). So I removed that line again…

Maybe there's another place to edit mount options in Ubuntu Mate? system.d? Elsewhere?

Sometimes (under Ubuntu MATE 16.04) I apparently press some evil shortcut, and all windows get minimized to the sidebar (Panel).

Strangely, I can not un-minimize (restore, maximize) them. Also opening a new Terminal (ctrl–T), that Terminal starts minimized. So cannot even restart the windows manager…

Clicking on them in the sidebar or plank, briefly shows the animation lines of opening/closing, perhaps showing the window for a millisecond, but that's about it…

All I can do in such situation is to reboot…

Any advice, what's wrong?

Switching to console and back (ctrl-F6/F7) does not help.

I am not talking "Hide all normal windows and set focus to the desktop" from the keyboard shortcuts. I have a shortcut for that, I am aware of it, and minimizing as well restoring with that does work. Must be something different…

Basically, Ubuntu (-Mate 16.04) plays sound output on the first non-disabled sound-device in the preferences list:

Only trouble: I'd like to change the order. If there's a headset or that USB audio device connected, that should come first, internal card only as the last resort. (Currently I have to disable all higher-up soundcards)

I played around with the pacmd set-default-sink command (Indices are 0,6,7 on my machine) but no luck...

Possibly related Question. — Possibly related Question..

I have two 300 GB files on different volumes:

• encrypted local backup
• encrypted ‘remote’ backup on NAS that is).

By design, these two files are identical in size and also mostly (>90%) identical in content...

Is there an efficient tool to „rsync“ these files, and only copy over the differing sections, so the target file becomes identical with the source?

Perhaps somethings that builds checksums of blocks to figure that out, I don't know... (anything more efficient than cp -f... rsync would afaik also grab the entire source file to overwrite)

I often want to hide my sidebar (from mate-panel's point of view it is called the bottom panel). Just to have more space. And unhide it, with a keyboard shortcut. So temporarily, not forever, to have more space and order on screen. How do I do that?

Yes, I am aware of this this this and this answer.

Still, I can't find the log information (the log lines I see when pressing ESC during splash screen. For example ‘stopped Target User and Group Name Lookups.’ ...) in any of the possible suspects in var/logs.

I.e. by looking at the most recently changed files: (my shutdown was at 11:20h, thus only these files can possibly matter)

/var/log $ ll -tr | tail
    -rw-r--r-- 1 root      root  27K 2016-10-27 11:03  Xorg.0.log.old
    -rw-rw-r-- 1 root      utmp    0 2016-10-27 11:20  wtmp
    -rw-r----- 1 syslog    adm     0 2016-10-27 11:20  kern.log
    -rw-r--r-- 1 root      root    0 2016-10-27 11:20  boot.log
    -rw-r--r-- 1 root      root    0 2016-10-27 11:20  gpu-manager
    -rw-r--r-- 1 root      root    0 2016-10-27 11:21  gpu-manager.log
    -rw-r----- 1 root      adm     0 2016-10-27 11:21  apport.log
    -rw-r--r-- 1 root      root    0 2016-10-27 11:21  Xorg.0.log
    -rw-r----- 1 syslog    adm  3,0K 2016-10-27 11:21  auth.log
    -rw-r----- 1 syslog    adm   194 2016-10-27 11:21  syslog

Did things change? Do I have to enable (persistent) shutdown logging somewhere?

My shutdown did not fail, took 15-20 seconds and completed. (Multiple runs before I asked this question, naturally). If you wonder about the 0-bytes: Truncated (emptied) those log files before shutdown for simplicity.