www.bernatchez.net - troubleshoot

Obscure Pelican Compilation Warning

2026-04-05T22:51:49+00:00

Obscure Warning Message

While performing a pelican compilation - "make html" - I got a warning which took me a long time to track down, like the one below.

WARNING  There are 2 original (not translated) items with slug "commonslugname":
         .../content/somedir/somename-fr.rst
         .../content/somedir/somename-en.rst

This type of warning usually means you forgot to change the :lang: attibute from en to es or to fr in the translated versions of the same .rst file.

This time, I checked the :lang: attributes on both files and they were correct. Everything else I looked at also seemed perfect.

It took me a long time to figure out, that the difference that was throwing off the compiler was that I had attibutes ":status: hidden" in both the -en, and the -fr, but not in the -es version.

So if you are making an article "hidden" be sure you do so in all translations.

Data Rescue

2025-10-29T16:57:39+00:00

Overview

I ran into trouble when I attempted to upgrade my laptop from ubuntu 18.04 to 20.04. My relesease_upgrade from Ubuntu 18.04 to Ubuntu 20.04 got hung up, so after waiting quite some time I decided to break out of the upgrade script.

I looked around and version 20.04 seemed well installed and working properly.

I hoped that the version would just work now, So I rebooted without taking any precautions.

The ubuntu boot sequence got hung up but I managed to use a rescue option in the Grub menu that booted an old archived partition of ubuntu16.04 that I happened to have on the drive.

I blindly assumed that my issue was merely which partition was set to active. So I changed that and rebooted and it made matters worse.

uhoh

output:

error: no such partition.
Entering rescue mode...  grub rescue>

Now nothing booted.

Doh!

What to do?

Look around.

grub rescue> ls

output:

(hd0) (hd0,msdos3) (hd0,msdos2)(hd0,msdos1)

Those are (device,partition) references.

ls (hd0)

output:

(hd0): Filesystem is unknown.

ls (hd0,msdos2)

output:

(hd0,msdos2): Filesystem is ext2.

Note It guessed ext2, but actually it is ext4.

ls (hd0,msdos2)/home

output:

./ ../ ubuntu/ pierre/

That is the partition I wanted

ls (hd0,msdos1)/home

output:

./ ../ dev/ sysadmin/

The partition that worked, now I want it back working.

Ubuntu live USB to the rescue.

Follow this article: Ubuntu Live Rescue

Ok Now we are back with booting the relic partition. Now this time Preserve Critical Info First

${HOME}/jnut

I maintain a subdirectory on my computer where I put essential files that are important to me. Files that I need to preserve from one incarnation of a workstation to another. Among other things, it contains my latest password database. I preserve this on USB flash drive.
Project sources which were being maintained on the faulty partition.

These projects are under revision control, off the workstation. We still need to preserve the information of where they are and how to access their git repostitory..
Anything else that needs preserving that is not already preserved on jnut.

To accomplish the preservation of the above I recursively copied "/mnt/ubuntu20.04/home/ubuntu/jnut" to USB flash drive.

I copied all the information for 2 and 3 above to a sub-directory "ship_toUSB" and I will also replicate that onto the flash drive. The file "ship_toUSB/billomaterials.txt" indicates what the directoy contains.

cat ship_toUSB/billomaterials.txt

output:

Bill of materials
=================

allsources.txt          List of project directories on the faulty partition.
billomaterials.txt      This file.
listrepos               Bash script to list repos existing on guanabana and give you a few tips.
mangobiche16_04.zip     Some old jpeg images I want to keep for personal reasons.
skeleton.kdbx           Transition password database.
venv.pdf                Doc helps with virtual environements.
vwrapper.pdf            Same as above.
venvlist                List of virtual environments under ubuntu on the faulty partition.
reqs/                   Contains requirements files corresponding to each item in venvlist.

Personal Reading lists:

  Norm-BookList-2021.docx.rtf
  Norm-BookList-2022.docx.rtf

Among those files two of them are informative enough to list them in this document.

cat ship_toUSB/allsources.txt

output:

# Our project sources make use of git repositories.
# for repos at gituser@guanabana.bernatchez.net you need the ssh private key: id_guanabana_admin
#     then git push/pull etc will work properly
# for repos at git@github.com:pierrebernatchez/ you need the key: mangobiche_github,
#
# Note: You can find both private keys in "jnut/private_pack/"
# and their passwords in the password database that holds them "jnut/pierreKeepass.kdbx"
# and also in "ship_toUSB/skeleton.kdbx"
#
# Working copies of these sources existed on the now faulty partition
# in these directories.
#
# from /home/ubuntu/collab:
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/wait_ssh.git
    url = git@github.com:pierrebernatchez/handyhelper.git
    url = git@github.com:promotionefx/animbboard.git
    url = git@github.com:pierrebernatchez/latexhelper.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/peripherals.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/restdocs.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/metarepos.git
    url = git@github.com:pierrebernatchez/mathsansmystere.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/personalnotes.git
    url = git@github.com:pierrebernatchez/msmsite.git
# from /home/ubuntu/repos:
    url = git@github.com:pierrebernatchez/srcgif.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/srctemplate.git
    url = git@github.com:pierrebernatchez/minifirewall.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/ubuntuprereq.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/bernatchez_staticweb.git
    url = git@github.com:pierrebernatchez/srcgeek.git
    url = git@github.com:pierrebernatchez/srcdrawhead.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/ogopogo_staticweb.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/geekutils.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/bboards.git
    url = gituser@guanabana.bernatchez.net:/home/gituser/repos/cssbootstrap.git

cat ship_toUSB/venvlist

output:

#
# List of virtual python environment directories that existed under
# the home/ubuntu/ directory on the faulty partition.
# I am not sure which if any were still relevant, nor which projects
# assumed which virtual environment.  But just in case, I have
# captured each one's context with a 'pip freeze' so that any one of
# these can presumably be resored for use with any any of the projects.
#
# The list indicates each environment's saved requirements context file.
#
devpi_venv             reqs/devpi_venv_requirements.txt
django_venv            reqs/django_venv_requirements.txt
latexhelper_venv       reqs/latexhelper_venv_requirements.txt
mathsansmystere_venv   reqs/mathsansmystere_venv_requirements.txt
msm_site_venv          reqs/msm_site_venv_requirements.txt
msmsite_venv           reqs/msmsite_venv_requirements.txt
restdocs_venv          reqs/restdocs_venv_requirements.txt

When I was ready to copy the files to the USB flash drive I ran into another issue. This resolved it.

USB drive issue

Epilog

Now this important data is backed up to USB flash drive I feel more adventurous.

I have decided to correct my 20.04 partition by simply re-using the partition space to install a recent version of UBUNTU onto it from scratch.

But first just as an added fallback precaution, I plan to deep copy everything under /home/ubuntu onto the old 16.04 partition.

Ubuntu Live Rescue

2025-10-29T15:22:02+00:00

Overview

How I Managed to return to booting from an old partition when I had changed the active boot partition using 'testdisk' and got myself into trouble.

Create/Fix Grub 2 boot

Boot from Ubuntu Live USB

Choose 'try ubuntu'. Boot Normal mode Without persistence Open a terminal window.
Detect Partition to install GRUB

sudo fdisk -l

output:

Disk /dev/sda: 465.8 GiB, 500107862016 bytes, 976773168 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disklabel type: dos
Disk identifier: 0x5f540829

Device Boot Start End Sectors Size Id Type
/dev/sda1 2048 156205055 156203008 74.5G 83 Linux
/dev/sda2 * 156205056 968665087 812460032 387.4G 83 Linux
/dev/sda3 968665088 976771071 8105984 3.9G 82 Linux swap / Solaris

Note:

/dev/sda1 is the one that we can still boot from.
/dev/sda2 is the one I made active (causing the problem)

Mount Partition(s)

sudo mount /dev/sda1 /mnt

Now point /dev and a few others to locations on the partition so grub uses those.

for i in /dev /dev/pts /proc /sys; do sudo mount -B ${i} /mnt${i} ; done`

Maskerade as though ''/mnt' were now ''/''

sudo chroot /mnt
# You can get back to the main root with <ctrl>d

install grub

grub-install --recheck /dev/sda
# The recheck checks/rebuilds the partition map grub uses.

update grub

update-grub
exit

unmount partitions

for i in /sys /proc /dev/pts /dev ; do sudo umount /mnt${i}; done

sudo umount /mnt

reboot computer

The steps above corrected my problem, but I still had to choose a grub recovery option to boot properly.

This is because the active partition is still set to the one that does not work. We need to switch the active partition.

To do that, I booted up using grub recovery option.

I Then logged into the Ubuntu 16.04 partition and used the GUI app called disks . I edited the (wrong) active partition to turn off "bootable", then edited the correct partition to turn it on.

After that reboot brought up the working partition hands free.

Note1

This Note is all but copied from the link where I learned how to do this stuff:

How To install/repair GRUB with Ubuntu Live CD/Flash

Note2

testdisk is a popular tool to recover data from faulty partitions.

But Since our partitions were pretty well intact, we did not need it in this case.

Automount Partition

2025-10-28T22:01:19+00:00

Overview

After booting up using an alternate partition I need to be able to access the partition that contains my latest files and information.

Since I don't want to have to mount manually each time I boot up I need to make an entry for the given partition in the "/etc/fstab" file so that it mounts automatically upon boot up.

Make a directory where we will mount the partition.

sudo mkdir /mnt/ubuntu20.04

This is in my /etc/fstab:

#
#
# Use 'blkid' to print the universally unique identifier for a
# device; this may be used with UUID= as a more robust way to name devices
# that works even if disks are added and removed. See fstab(5).
#
# <file system> <mount point>   <type>  <options>       <dump>  <pass>

UUID=f90fb201-6ba5-4bff-aad0-a127d20e1a63 /               ext4    errors=remount-ro 0       1
UUID=e9c9590f-c1f9-49f0-a83f-820c64728433 none            swap    sw                0       0

We want to append a line there for our as yet unmounted partition.

One that looks like this:

UUID=nnnnnnnn-nnnn-nnnn-nnnn-nnnnnnnnnnnn /mnt/ubuntu20.04 ext4 errors=remount-ro 0 1

Find out the information we need.

sudo fdisk -l

output:

Disk /dev/sda: 465.8 GiB, 500107862016 bytes, 976773168 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disklabel type: dos
Disk identifier: 0x5f540829

Device Boot Start End Sectors Size Id Type
/dev/sda1 * 2048 156205055 156203008 74.5G 83 Linux
/dev/sda2 156205056 968665087 812460032 387.4G 83 Linux
/dev/sda3 968665088 976771071 8105984 3.9G 82 Linux swap / Solaris

We can see that the partition we want is the one at /dev/sda2. Discover the UUID of the partition we want to mount

sudo blkid

output:

/dev/sda1: UUID="f90fb201-6ba5-4bff-aad0-a127d20e1a63" TYPE="ext4" PARTUUID="5f540829-01"
/dev/sda2: UUID="8219c564-8865-44c8-a691-15faa4d3db4f" TYPE="ext4" PARTUUID="5f540829-02"
/dev/sda3: UUID="e9c9590f-c1f9-49f0-a83f-820c64728433" TYPE="swap" PARTUUID="5f540829-03"
/dev/loop0: TYPE="squashfs"
/dev/loop1: TYPE="squashfs"
/dev/loop2: TYPE="squashfs"
/dev/loop3: TYPE="squashfs"
/dev/loop4: TYPE="squashfs"
/dev/loop5: TYPE="squashfs"

So the line we need to append to /etc/fstab is this:

UUID="8219c564-8865-44c8-a691-15faa4d3db4f" /mnt/ubuntu20.04 ext4 errors=remount-ro 0 1

Append a comment line then our automount line to /etc/fstab.

Beware here, we are modifying an important file, don't damage the file and cause a boot up issue. Use your UUID not the one in echo example below.

echo '# Appended Automount partition at /mnt/ubuntu20.04' | sudo tee -a /etc/fstab`
echo 'UUID=8219c564-8865-44c8-a691-15faa4d3db4f /mnt/ubuntu20.04 ext4 errors=remount-ro 0 1' | sudo tee -a /etc/fstab

Double check:

cat /etc/fstab

# /etc/fstab: static file system information.
# Use 'blkid' to print the universally unique identifier for a
# device; this may be used with UUID= as a more robust way to name devices
# that works even if disks are added and removed. See fstab(5).
#
#
# / was on /dev/sdb1 during installation

output:

UUID=f90fb201-6ba5-4bff-aad0-a127d20e1a63 / ext4 errors=remount-ro 0 1
# swap was on /dev/sdb5 during installation
UUID=e9c9590f-c1f9-49f0-a83f-820c64728433 none swap sw 0 0
# Appended Automount partition at /mnt/ubuntu20.04
UUID=8219c564-8865-44c8-a691-15faa4d3db4f /mnt/ubuntu20.04 ext4 errors=remount-ro 0 1

Retrigger mounts.

sudo mount -a

OK Job is done. Make sure your mounts are ok before rebooting.

Issues with Flameshot utility ubuntu 24.04

2025-10-28T18:11:40+00:00

In my opintion the best screenshot utility availabe under ubuntu is flameshot.

To install:

sudo apt install flameshot

But under ubuntu 24.04 there is a problem with the installed package.

When you try to take a screenshot you get this message:

Flameshot Error
Unable to capture screen

The problem is that the package assumes XORG rather than wayland, for its interaction with the display. So one needs to get it to assume wayland instead.

Some people have been using an autostart script which invokes flameshot with the correct parameter. The consequence is a flameshot mini-icon which when used behaves correctly.

To get that you just need to create an autostart desktop here:

${HOME}/.config/autostart/flameshot-custom.desktop

With these contents:

[Desktop Entry]
Type=Application
Name=Flameshot wayland fixed
Exec=sh -c "QT_QPA_PLATFORM=wayland flameshot"
X-GNOME-Autostart-enabled=true

But I prefer to use a keyboard shortcut rather such a mini icon to get my screenshots.

You set that up by using the settings app to add a shortcut:

settings->keyboard->view and customize shortcuts->custom shortcuts->+

Name: Flameshot

Command:

sh -c "QT_QPA_PLATFORM=wayland flameshot gui"

Shortcut: You choose the key you want.

I like to use printscreen.

On my system printscreen is <fn><right-shift>

Using that overrides the default screenshot which also uses printscreen.

Loading pre-existing SSH key to seahorse

2025-10-27T17:32:31+00:00

I have been accustomed to just using ssh-add by hand whenever I access remote ssh sites.

But now that I want to start fully utilizing PGP keys for email privacy too, I felt I should use one comprehensive tool to manage all the keys I am using.

I tried using the "passwords and keys" (seahorse) user interface to add an existing SSH keypair, but there was no way to do that in the user interface, you can only add by generating new keys.

However it turns out that seahorse keeps in touch with ssh agent keys.

So all you need to do is to load your key into the agent and seahorse will take it in.

For the sake of keeping things in order copy the pair to .ssh/

cp pathto/my_sshkey* ${HOME}/.ssh/

Load the key

ssh-add .ssh/my_sshkey
## you need to supply the passphrase for the private key

Make sure it got loaded

ssh-add -L

Now if you run seahorse you find it is there.

SSH connection refused

2025-10-27T15:52:30+00:00

On my recently installed Ubuntu 24.04 laptop.

When I attempted to ssh into my git server my ssh connection was refused.

I checked for all the usual errors and the problem persisted.

This version of ubuntu uses a key management tool called "passwords and keys".

To get that to play well with my keys, the course of least resistance was to remember that it likes to find ssh keys in ${HOME}/.ssh/ and it likes to find both private and private key files in there.

When I conform to that restriction ssh behaves correctly with my github account repository.

This shows any such keys in its list.

ssh-add -l

However trying to connect to my own repo server still yielded: connection refused "public key"

The measure that fixed that issue was to add two lines to the file: /etc/ssh/ssh_config.

Immediately below the "Host *" line.

Host *
    PubkeyAcceptedKeyTypes=+ssh-rsa
    HostKeyAlgorithms=+ssh-rsa

USB mount error with exfat file system

2025-10-26T13:56:13+00:00

mount error: unknown filesystem type 'exfat'

When I inserted a USB flash drive it seemed to mount properly, but when I tried to browse there I got the the above message.

To fix the problem I had to install the appropriate exfat software:

sudo apt update
sudo apt install exfat-fuse exfat-utils