No luck with 4.9.0-3, either. I tried most of the kernels available in ascii's main and backports repositories, and all of them give me more or less severe video artifacts.

I decided to stop fighting this and simply fetched and installed the kernel I was using on devuan jessie (linux-image-3.16.0-5-amd64). Lo and behold, no more video glitches. Woo-hoo!

Gvfs-mtp-volume-monitor seems to be a buggy mess. There only consistent thing about it is that it always detects my phone twice. Otherwise, it is pretty unpredictable: It occasionally detects the phone and shows the two desktop icons, but doesn't actually mount the device. Occasionally the caja windows that pop up do not contain any files or directories, so I need to close the two windows then double-click on one of the two desktop icons. Occasionally, the device is detected but an error box pops up saying the device is unavailable. I give up on using gvfs for mounting my phone.

Here is my graybearded workaround:

1. sudo apt-get install jmtpfs

2. sudo chmod a-x /usr/lib/gvfs/gvfs-mtp-volume-monitor

3. Create /etc/udev/rules.d/80-mtp-mount.rules with this in it:

ACTION=="add", SUBSYSTEM=="usb", ATTR{idVendor}=="04e8", RUN+="/bin/su bruno -c /opt/scripts/mtp-mount"

4. Create /opt/scripts/mtp-mount with this in it, then make it executable:

#!/bin/sh

# mount the device
export PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/opt/scripts
export DISPLAY=:0.0
mkdir $HOME/Phone && jmtpfs $HOME/Phone && caja $HOME/Phone

# create a .desktop file to unmount the device
desktop_file=$HOME/Desktop/Phone-unmounter.desktop
cat << EOF > $desktop_file
[Desktop Entry]
Comment=Phone Unmounter
Exec=sh -c "fusermount -u ~/Phone; rm -r ~/Phone; rm $desktop_file"
Icon=media-eject
Terminal=false
Type=Application
Name[en_US]=Eject Phone
EOF
sleep 0.5 # to make sure .desktop exists
sudo chmod a+x $desktop_file

The above is as reliable as a faithful old dog, with never any surprises.

Folks experiencing flaky mounting of their Android phones (or other mtp device) can recycle the above workaround with only minimal modifications:
  - change "bruno" to your username in the udev rule
  - change the idVendor attribute in the udev rule to match your device's manufacturer

Devuan Jessie and an updated Arch Linux only show the phone once when it's plugged in.

I tried re-installing all packages with gvfs in the name. Also, in caja's media preferences I unchecked "Browse media when inserted" and checked "Never prompt or start programs on media insertion". Still, two desktop icons show up on the desktop when the phone is connected--the only difference now is that two caja windows do not automatically pop up when I plug in the phone.

Therefore, the issue here is not that caja is detecting the device twice: gvfs is the one that's detecting the device twice. Alas, I think I have the right diagnosis but still have no clue regarding the proper treatment.

That's a good thought, but it only has an internal/simulated sd card. Also, Devuan Jessie only shows it once.

I have an old Arch Linux machine I've kept updated. Will try plugging into that and will update this thread with result.

I found a workaround: An older version of yad (0.27.0-1.1 instead of the 0.38.2-1 available in ascii repository) handles the taskbar icons perfectly.

The elevated credentials icon is still messed up, but I see that icon 1/10 of the time I see various icons placed in taskbar by yad, so I am happy with the workaround.

My gut feeling is that the buggy taskbar icons are due to a GTK3 version incompatibility between current yad/gnome-keyring and current mate-panel, but I wouldn't know how to test this hypothesis. If someone knowledgeable knows the actual reason and/or a proper fix, please do share.

All machines in my household run devuan and receive unattended security upgrades. Would you happen to know whether non-default kernels receive security upgrades through the security repository?

I'd like to have my cake and eat it too, of course, but if I *have* to choose between automatic security upgrades and fixing a minor cosmetic issue then I'll take the upgrades.

Thank you, Geoff42. I hadn't tried lxqt-powermanagement. I just did and it is very nice. Thanks for the recommendation.

No luck. I purged all three live-config packages and ancron, rebooted, and reinstalled anacron. Similar shenanigans (at least now the symlink is pointing to something sensible and not /bin/true):

bruno@thinkpad:~$ ls -l /usr/sbin/anacron
lrwxrwxrwx 1 root root 30 Feb 24 01:36 /usr/sbin/anacron -> /usr/sbin/anacron.orig.anacron
bruno@thinkpad:~$ dpkg -S anacron.orig.anacron
diversion by live-config from: /usr/sbin/anacron
diversion by live-config to: /usr/sbin/anacron.orig.anacron

This is my /etc/apt/sources.list:

deb http://pkgmaster.devuan.org/merged/ ascii main 
deb http://pkgmaster.devuan.org/merged/ ascii-security main
deb http://pkgmaster.devuan.org/merged/ ascii-updates main

fsmithred, are you sure that your /usr/sbin/anacron is a real file and not a link? If you are sure, what am I missing?

What is this live-config business anyway? Can I get rid of it?

After much snooping around, I discovered that /usr/sbin/anacron was a link to /bin/true (???)
The anacron package actually installs /usr/sbin/anacron.orig.anacron (???)
This fixed the issue: sudo ln -fs /usr/sbin/anacron.orig.anacron /usr/sbin/anacron

Very, very strange.