a-d.s - wiki

CHIRP

2025-06-13T00:00:00+02:00

Personal settings
Finding which port (Arch Linux)
"Radio did not respond"

Personal settings#

These are settings for the radios and setup I use. It's probably not very useful to anyone else.

port: USB serial (/dev/)ttyUSB0
Radio model: UV-S9X3

Finding which port (Arch Linux)#

To find on which port the radio is connected, open a terminal and type journalctl -f -n 50, to see system log as they happen. Connect the cable, and look at the logs. You should have something like this:

1 Jun 13 01:35:41 niflheim kernel: usb 1-1: new full-speed USB device number 16 using xhci_hcd
2 Jun 13 01:35:42 niflheim kernel: usb 1-1: New USB device found, idVendor=1a86, idProduct=7523, bcdDevice= 2.64
3 Jun 13 01:35:42 niflheim kernel: usb 1-1: New USB device strings: Mfr=0, Product=2, SerialNumber=0
4 Jun 13 01:35:42 niflheim kernel: usb 1-1: Product: USB Serial
5 Jun 13 01:35:42 niflheim kernel: ch341 1-1:1.0: ch341-uart converter detected
6 Jun 13 01:35:42 niflheim kernel: usb 1-1: ch341-uart converter now attached to ttyUSB0

This last line [...] attached to ttyUSB0 is what you're looking for. Either select that option in "port", or select "custom" and enter /dev/ttyUSB0

"Radio did not respond"#

Aside from typical troubleshooting (e.g. cable not seated properly, as it requires way more force than you may think), wrong selection of port can also trigger a Radio did not respond error.

Wordpress

2025-06-10T00:00:00+02:00

Content

Security
- disable xmlrpc.php

Security#

disable xmlrpc.php#

https://www.hostinger.com/fr/tutoriels/xmlrpc-php

Games

2024-08-06T00:00:00+02:00

Content

Minecraft
BallisticNG
- reset progression & cloud
KSP
- Launch command

Minecraft#

Good tutorial on how to install Minecraft server on Linux: https://www.shells.com/l/en-US/tutorial/0-A-Guide-to-Installing-a-Minecraft-Server-on-Linux-Ubuntu

BallisticNG#

reset progression & cloud#

Launch the game, go back to steam without closing it, right click > browse local files, delete "save" all while it runs, it should do the trick

KSP#

Launch command#

Fixing various control and localization bugs, plus dGPU launch

DRI_PRIME=1 SDL_DYNAMIC_API=/usr/lib64/libSDL2.so LC_ALL=C  %command%

Godot

2023-10-24T00:00:00+02:00

Content

Multimesh buffer's layout
MultiMeshInstance (2D)

I usually work in csharp but most of the code here can be easily translated into godot's language "GDScript" if necessary using the API's guide.

Multimesh buffer's layout#

The Multimesh property of a MultiMeshInstance3D node exposes its raw data through a Buffer property allowing for instance to apply updated data from a compute shader directly.

Its layout is however fairly opaque and hidden in other wiki pages and in the source code: so here is a snippet to avoid (my) hours of headbanging research.

Each instance has a Transform3D property describing its position, rotate and scale, itself composed of an Vector3 Origin describing its position, and a 3x Vector3 Basis describing its rotation and scale. Each Vector3 is being made of three floats values, forming a set of 12 float values in a (in my opinion) non-intuitive order.

If you enable UseColors=true, four additional values (R, G, B and A, as a Color property) are added at the end of this set.

If you enable UseCustomData=true, four additional values (for well, custom usages) are added at the end of this set.

Depending on the configuration, your data set will be of size 12, 16 or 20, and you need to offset you id's by that factor.

 1 int offset 16; //3D base + color enabled, no custom data
 2 int BufferId = offset * id;
 3 
 4 data[BufferId + 0] = value.Basis.X.X;
 5 data[BufferId + 4] = value.Basis.X.Y;
 6 data[BufferId + 8] = value.Basis.X.Z;
 7 
 8 data[BufferId + 1] = value.Basis.Y.X;
 9 data[BufferId + 5] = value.Basis.Y.Y;
10 data[BufferId + 9] = value.Basis.Y.Z;
11 
12 data[BufferId + 2] = value.Basis.Z.X;
13 data[BufferId + 6] = value.Basis.Z.Y;
14 data[BufferId +10] = value.Basis.Z.Z;
15 
16 data[BufferId + 3] = value.Origin.X;
17 data[BufferId + 7] = value.Origin.Y;
18 data[BufferId +11] = value.Origin.Z;
19 
20 data[BufferId +12] = value.Color.R;
21 data[BufferId +13] = value.Color.G;
22 data[BufferId +13] = value.Color.B;
23 data[BufferId +13] = value.Color.A;

MultiMeshInstance (2D)#

In the case of 2D the same applies, but with Origin being only a Vector2 and Basis being only 2x Vector2, forming a 2+2x2[+ 4][+ 4] data set of size going from 6 to 14.

Krita

2023-08-12T00:00:00+02:00

Content

Brushes
multiple columns for tools
Hide title in canvas mode

Brushes#

Krita brushes 2023-01 bundle

multiple columns for tools#

(eg: for tablet mode left-handed)

first put dock next to other one (even if it doesn’t take whole height), then take the dock below and put it fully under.

see: https://www.youtube.com/watch?t=161&v=uYdEkOyFUn8&feature=youtu.be

Hide title in canvas mode#

in settings, switch multiple documents from “tabs” to “subwindow”

see: https://krita-artists.org/t/hide-top-bar/13076/2

Thunderbolt 3 & eGPU

2023-07-10T00:00:00+02:00

Content

thunderbolt
eGPU
Blender renderer
External screen setup

This is an old and chaotic list of tips: it's probably no longer relevant

thunderbolt#

https://forum.manjaro.org/t/cant-get-extra-monitor-interface-to-show-up/28767/5

blender 3.5

HSA_OVERRIDE_GFX_VERSION=10.3.0 DRI_PRIME=1 ./blender

eGPU#

https://wiki.archlinux.org/title/PRIME

glxinfo | grep "OpenGL renderer"

DRI_PRIME=1 glxinfo | grep "OpenGL renderer"

Blender renderer#

https://developer.blender.org/T102018 (5.4 is equivalent of 22.xx)

https://aur.archlinux.org/packages/ncurses5-compat-libs with gpg --keyserver keyserver.ubuntu.com <http://keyserver.ubuntu.com/) --recv-keys 19882D92DDA4C400C22C0D56CC2AF4472167BE03
opencl-amd https://bbs.archlinux.org/viewtopic.php?id=268346

LoadLib(libhsa-amd-aqlprofile64.so) failed: libhsa-amd-aqlprofile64.so: cannot open shared object file: No such file or directory → sudo pacman -S arch4edu/hsa-amd-aqlprofile-bin

How to Install RPM Packages on Arch Linux — marcelo cubillos

90 degree 8 pin connector - Google Search

Support for RX6600XT (gfx1032,navi23)? · Issue #1698 · RadeonOpenCompute/ROCm · GitHub

External screen setup#

with all-ways-egpu, enable only option 1, reboot, then sudo all-ways-egpu set-compositor-primary egpu`

Virtual machines on Linux

2022-06-03T12:40:00+02:00

Content

Installation of the gaming machine
Moving a VM

Installation of the gaming machine#

TODO when reinstalling.

Moving a VM#

Based on `Moving KVM/QEMU VMs <https://kamprianis.eu/michalis/i.think/personal/200628-moving-kvmqemu-vms-.html>`__ on kamprianis.eu

Shut VM down
Use virsh list --all, sudo virsh list --all and sudo virsh pool-list --all respectively to get infos
sudo mv <old_path>/mv.qcow2 <new_path>/mv.qcow2
sudo service libvirtd restart
sudo virsh vol-list to check if machine is recognized
sudo virsh edit debian10, search for old path and replace it.
boot VM to check if all is well.

Basic linux packages

2022-02-03T18:30:00+01:00

1    sudo apt install vim sysstat mlocate tmux python3-pip htop rofi xbindkeys keepassxc rsync fd-find hwinfo gparted curl

Homebrew cloud gaming setup

2022-02-03T18:30:00+01:00

Content

Installation of the gaming machine
Turning the machine on and off remotely
- Turning on
- Turning off
Controlling from a smartphone (yes!)
Troubleshooting

Installation of the gaming machine#

If you want to game on linux, see Linux gaming setup for a smooth Linux install.

Turning the machine on and off remotely#

This is an optional step, as you can just leave your machine running 24/7. But if you care about electricity bill and your hardware, this is how to turn on and off your machine.

Turning on#

Use local server to wake up machine (require setting up Wake on LAN in gaming machine's BIOS/UEFI)

You can use a SSH script and SSH key authentification to have a simple "turn on" shortcut on your remote machine running linux:

nemecle@niflheim > ~ > cat wol_ultimatebob.sh
#!/bin/bash set -Eeuo pipefail

ssh nemecle@server.local -p 34423 wakeonlan <mac_address>

Turning off#

From the Smartphone Steam link, simply use the dedicated button

From a computer, add a non-steam game on the gaming computer, select any program, change the shortcut properties to "C:\Windows\System32\shutdown.exe" -s -t 00, or /sbin/poweroff if it runs on linux, and call that "game" remote shutdown. When you want to turn off the computer remotely, simply "stream" the game.

More info

Controlling from a smartphone (yes!)#

It's actually possible to control the computer from a smartphone. Even over LTE, it is still playable, as long as you are confortable with using a controller rather than a keyboard-mouse setup.

Simply download the Steam Link (more info and other devices here), Go to settings ⚙️, Computer, hit Other Computer. Write down the PIN.
On the computer, open Steam, go to Settings, Remote Play, and PAIR STEAM LINK, and enter the PIN.
Get yourself a smartphone controller. Touch control is a sham.

Tadaaaaa!

Borderlands 3, running on debian. What an era.

Troubleshooting#

If Steam shows "install" on remote machine despite game being installed on gaming machine, select remote machine with down arrow next to it and click "install" anyway: it will launch game as stream anyway

On a windows computer, Steam sometimes start as delayed, or doesn't start at all, when the computer is remotely shut down. It seems like the shutdown command doesn't actually behaves like the shutdown button.

Linux gaming setup

2022-02-03T18:30:00+01:00

Content

Installation of the gaming machine
Remote gaming

Installation of the gaming machine#

For some reasons, typical "user-friendly" distributions are a shit circus. People usually system wide bugs with drivers that I don't with my own setup, so as I reinstalled my computer I took notes of what I did. It seems that beside the lack of necessary proprietary drivers for some features, Debian is simpler to use, even for gaming.

Debian might seem like an odd choice for two reasons: for starter, it lags behind software updates (except security updates) to only use well tested software. This can be "solved" by using the Sid unstable version, which is actually still quite stable and can be used for daily driving. Also, debian doesn't integrate proprietary software by default, so the corresponding repositories must be activated for Steam and Nvidia, even if it goes against debian philosophy.

I believe that any "base" linux distribution might also offer the same benefits without these limitations, but I'll let someone else test that hypothese.

Long story short, so called "end user friendly" distributions seem to actually be worse, and plenty of "base" distributions offer similar user friendly ecosystems [1].

[1]	Even though they also suffer from the usual "xxxx Is the Year of Linux on the Desktop!!!" syndrome.

1: . Install Debian with Gnome or MATE. If you're not used to installing Linux, get in touch with free Open-Source software organisation near you or go to an Install party dedicated for non-power users. 2: Enable non-free AND contrib repositories. Don't forget the last one or you'll debug missing packages for nothing (learn that the hard way). 3: Install Steam. It requires some commands, but nothing to scary. Just enter those lines in the terminal.

1 sudo dpkg --add-architecture i386
2 sudo apt update

And install steam using the software provided by Gnome or MATE, or use this command:

1 sudo apt install steam

More info here.

4: Install nvidia drivers and various utilities

1 nvidia-drivers mesa-vulkan-drivers libglx-mesa0:i386 mesa-vulkan-drivers:i386 libgl1-mesa-dri:i386

Or once again, use the command line:

1 sudo apt install nvidia-drivers mesa-vulkan-drivers libglx-mesa0:i386 mesa-vulkan-drivers:i386 libgl1-mesa-dri:i386

Open Steam, log in, and then go to Settings > Steam Play > Enable Steam Play for all other titles

Now, simply use Steam as usual.

Games should then actually work according to ProtonDB rating.

Remote gaming#

See Homebrew cloud gaming setup

Self-hosted Spotify

2021-05-19T18:30:00+02:00

The is a work in progress: this page is here to keep track of the project.

The go-to reference in the self-hosting commnunity is very complex: https://www.reddit.com/r/selfhosted/comments/9ya0or/selfhosted_music_playerspotify/ea0m2bk/

The goal is to build a simpler version of it.

Nextcloud Install "Music" App

setup last.fm api key for informations

Generate login-pwd for Ampache/subsonic

The logins for subsonic are not the user logins, but the generated login-pwd pair from the Music app

use "Subsonic" android app for streaming from Nextcloud

Use last.fm android app to scrobble listening music

Anything below is TODO

Build Headphones (https://github.com/rembo10/headphones) -> Transmission -> beet -> Nextcloud Music pipeline

Ansible

2021-05-17T18:30:00+02:00

Content

On client
ad-hoc command
On server

On client#

1 sudo apt update && sudo apt install ansible -t
2 git clone https://framagit.org/Nemecle/heimdall
3 cd heimdall

to launch:

1 ansible-playbook -i inventory.ini site.yml

ad-hoc command#

1 ansible  -i inventory.ini webservers -m ansible.builtin.file -a "dest=/etc/nginx/sites-available/wiki.nemecl.eu mode=600 owner=www-data group=www-data"  --become

On server#

1 sudo adduser deploy
2 sudo groupadd wheel
3 sudo usermod -aG sudo deploy
4 sudo usermod -aG wheel deploy

Then sudo visudo, add %wheel ALL=(ALL) NOPASSWD: ALL

And edit /etc/shadow to add/edit deploy:!!:[...] for no password

make sure to add user to AllowUsers in sshd_config.

Then:

1 sudo service ssh reload

Add ssh key:

1 ssh-copy-id -p <port> -i $HOME/.ssh/id_rsa.pub deploy@nemecl.eu

(unecessary?) then:

1 sudo setfacl -R -m u:deploy:rwx /var/www/

https://linuxbuz.com/linuxhowto/what-is-ansible-playbook-and-how-it-works

Software

2021-05-12T22:00:00+02:00

Content

Learning/Education
- music
code
Graphics
learn
fun

Learning/Education#

music#

Incredibox
Beatmaker
Blokdust
Sampulator
musiclab
Drum Machine
Collaborative music: Plink
List of various sound editors https://github.com/everestpipkin/tools-list#sound-editors

code#

basic algorithms with shapes: Algoblocks
actual python or javascript code: CodeCombat
visual programming: Cables

Graphics#

For 2D animation and even interactive books: Wick Editor

learn#

morse: https://morse.withgoogle.com/learn/
keyboard "touch typing" : https://zty.pe/
mouse agility: https://kbhgames.com/game/ninja-glove

fun#

3D/VR chat in the browser: Hub by Mozilla
Experiment with alchemy/cellular automata: Sandspield
3D puzzles Rotopo

Move user folders in Windows

2021-05-12T18:30:00+02:00

The screenshot here are in french, but the corresponding translations are provided.

Following the configuration of new computers with an additional "Data" drive, I wanted to move users' folders (Downloads, Documents, Pictures, Videos,...) on it to keep the "system" disk for applications. Several articles and forum posts online advise against moving the whole user folder because it can cause problems with updates and such.

Also, most articles simply explain how to move it, but do not explain how to "recreate" the users' folder structure (which can actually be handled by Windows itself). Here is the quickest method I found without Active Directory and such.

To start with, connect with one of the users, and open the file browser. Under "Quick access"/"Shortcuts", right click on the corresponding shortcut, and click "Properties":

Then, go to the "location" tab. You should have something that look like this, with the default location being C:\users\<user>\Documents:

Simply replace the C: by D::

A pop-up appears, warning you that the targeted folder does not exist. the first "trick" being that it suggests to create it for you! Click on "yes" to create it:

A second pop-up will ask if you want to move the content of the old folder to the new: to which you can obviously answer "yes":

Apply this to all the folders that you want.

The hierarchy and the folders are automatically created, and their content seemlessly moved:

File recovery

2021-05-11T18:30:00+02:00

to redact

basic extract: testdisk (good tutorial: https://itsfoss.com/recover-deleted-files-linux/)

mp4fixer:

1 perl /home/nemecle/Downloads/mp4fixer/fixer.pl /media/nemecle/HDD2/b/testdisk/DCIM/101_FUJI/DSCF0658.MOV /media/nemecle/HDD2/b/testdisk/DCIM/101_FUJI/video.MOV video.MOV_fixed;

data carving: https://github.com/sleuthkit/scalpel https://www.cgsecurity.org/wiki/PhotoRec#How_PhotoRec_works http://foremost.sourceforge.net/

atom mov https://video.stackexchange.com/questions/29073/need-help-to-fix-a-broken-mp4-which-lacks-a-moov-atom https://codecpack.co/download/recover-mp4.html https://stackoverflow.com/questions/35561846/recovering-corrupted-video-file

data carving: https://github.com/sleuthkit/scalpel https://www.cgsecurity.org/wiki/PhotoRec#How_PhotoRec_works http://foremost.sourceforge.net/

Hardware

2021-05-11T18:30:00+02:00

Content

Storage
- Stuck head
- Bricked SSD
Connectors
- Power supply cords
- In-computer power connectors
Pc building
Seemingly dead computer

Storage#

Stuck head#

Diagnosis: "tick tick tick" and/or screeching noises when starting drive.

Either hit drive against a table (while unpowered) to unstuck: only do that with already fucked-up-to-be-backed-up drives.

(When I was a sysadmin, I heard a story of how old VAX drives would stall, so holes had been drilled in them with tape over the holes. When stalled, the sysadmin would peel back the tape and use their finger to spin-start them. Those even older drives must have been more tolerant of dust!)

More than once I had a hard drive fail to start up after a power cycle (back then the drives only spun down when power was removed). First thing we tried was to remove the drive and give the whole drive a sharp spin on the axis of the platter. Due to inertia of the platter, this would tend to get the platter to move a bit and "unstick" it.

My recollection is that it worked every time I had to do this. Of course, we would back up that drive and replace it as soon as possible.

Source

Bricked SSD#

Bricked SSD on Life Support (archive)

Connectors#

Power supply cords#

On power unit, ampere is a maximum (an equipement using less will be fine), while voltage is an exact value, with a tolerance of ≈10%

Plug "shape" is defined by IEC 60320

IEC 60320 norm as shown on Wikipedia

In-computer power connectors#

Power and Panel Cables and Connectors (archive)

Power supply labels on (mainly old) PSU connectors seems to have very little documentation. From here and here (archive), it seems to be:

label	Corresponding equipement
P1	System board, 24-pin
P2	Diskette drive
P3	CPU power, 6-pin
P4	1st SATA hard drive
P5	2nd SATA hard drive
P7	Optical drive

Pc building#

PC Part picker

Seemingly dead computer#

If a computer no longer powers (typically, a laptop), unplug power supply, remove battery, then press and release power button for one second, 30 times, and put battery and power supply back. It will discharge static charges from capacitors, actually removing any leftover power, and might allow computer to come back online.

The Secret Thinkpad Powerbutton Code To Bring Dead Laptops Back To Life (Hacker News thread with more tips here)

ROracle installation

2021-05-11T18:30:00+02:00

ROracle rough installation bash history

 1 1008  [12/01/18 - 16:30:24] R CMD INSTALL ROracle_1.3-1.tar.gz
 2 1009  [12/01/18 - 16:33:07] yum list oracle
 3 1010  [12/01/18 - 16:33:16] yum search oracle
 4 1011  [12/01/18 - 16:33:50] yum install oracle-instantclient12.1-devel.x86_64
 5 1012  [12/01/18 - 16:34:06] locate oracle
 6 1013  [12/01/18 - 16:34:26] find /usr -name "oracle"
 7 1014  [12/01/18 - 16:34:46] R CMD INSTALL --configure-args='--with-oci-lib=/usr/lib/oracle/12.1/client64/lib --with-oci-inc=/usr/include/oracle/12.1/client64' ROracle_1.3-1.tar.gz
 8 1015  [12/01/18 - 16:35:33] stat /softwares/R/library/ROracle/libs/ROracle.so'
 9 1016  [12/01/18 - 16:35:35] stat /softwares/R/library/ROracle/libs/ROracle.so
10 1017  [12/01/18 - 16:39:27] stat =/opt/oracle/instantclient
11 1018  [12/01/18 - 16:39:30] stat /opt/oracle/instantclient
12 1019  [12/01/18 - 16:39:56] locate oracle
13 1020  [12/01/18 - 16:40:31] export LD_LIBRARY_PATH=/usr/lib/oracle/12.1/client64:$LD_LIBRARY_PATH
14 1021  [12/01/18 - 16:40:34] R CMD INSTALL --configure-args='--with-oci-lib=/usr/lib/oracle/12.1/client64/lib --with-oci-inc=/usr/include/oracle/12.1/client64' ROracle_1.3-1.tar.gz
15 1022  [12/01/18 - 16:40:48] locate libclntsh.so.12.1
16 1023  [12/01/18 - 16:41:01] export LD_LIBRARY_PATH=/usr/lib/oracle/12.1/client64/lib:$LD_LIBRARY_PATH
17 1024  [12/01/18 - 16:41:02] R CMD INSTALL --configure-args='--with-oci-lib=/usr/lib/oracle/12.1/client64/lib --with-oci-inc=/usr/include/oracle/12.1/client64' ROracle_1.3-1.tar.gz
18 1025  [12/01/18 - 16:41:13] R
19 1026  [12/01/18 - 16:43:46] history

CSS only masonry layout

2021-04-23T19:14:00+02:00

Content

What is a masonry layout
Code

What is a masonry layout#

A masonry layout is simply a gallery of images in which elements are scaled to obtained a single block with straight borders on the left and the right (in the case of an horizontal layout).

While doing a basic layout is relatively easy, it gets harder when the direct children of the gallery aren't the images themselves, for instance when they are links to the page containing its full size.

Example:

Code#

html for the previous example:

 1 <div class = "flex-masonry">
 2   <a href ="/images/el-lissitzky_beat-the-whites-with-the-red-wedge.png">
 3    <img src = "/images/el-lissitzky_beat-the-whites-with-the-red-wedge.png" alt = "el lissitzky beat the whites with the red-wedge">
 4   </a>
 5 
 6   <a href ="http://www.phrack.org/issues/61/1.html">
 7     <img src = "/images/phrack61_intro.png" alt = "phrack61 zine intro">
 8   </a>
 9 
10    <a href ="/images/weingart.png">
11     <img src = "/images/weingart.png" alt = "weingart">
12   </a>
13 
14    <a href = "https://brandur.org/minimalism">
15     <img src = "/images/brandur.org.png">
16   </a>
17 
18    <a href = "http://cachestocaches.com/2018/9/ai-translation-more-language/">
19     <img src = "/images/cachestocaches.com.png">
20   </a>
21 </div>

CSS:

 1  .flex-masonry {
 2    display: flex;
 3    flex-wrap: wrap;
 4    margin: auto;
 5  }
 6 
 7  .flex-masonry a {
 8    overflow: hidden;
 9  }
10 
11  .flex-masonry a img {
12    max-height: 20vh;
13    object-fit: cover;
14    transform: scale(1);
15    transition: all 0.3s ease;
16    width: auto;
17  }
18 
19  .flex-masonry a img:hover {
20    transform: scale(1.05);
21   }
22 
23  /*https://medium.com/@iamryanyu/how-to-align-last-flex-item-to-right-73512e4e5912*/
24  .flex-masonry:last-child {
25    margin-left: auto;
26  }
27 
28  .flex-masonry > a {
29    align-items: stretch;
30    display: flex;
31    flex-direction: column;
32    flex: auto;
33    margin: 1px;
34  }
35 
36 
37  @media only screen and (max-width: 800px) {
38    /* For mobile phones: */
39    /* switch to one image per line, and no zoom effect */
40    .flex-masonry a {
41      max-width: none;
42      display: initial;
43    }
44 
45    .flex-masonry a img {
46      max-height: initial;
47      min-height: initial;
48      height: auto;
49      width: 100%;
50    }
51 
52    .flex-masonry a:hover img {
53      transform: none;
54    }
55  }

strace

2021-04-23T19:05:00+02:00

Content

What strace is
Basic commands
Filtering
Reading the file(s)

What strace is#

A quick excerpt from the man page to learn what strace is:

``strace`` is a useful diagnostic, instructional, and debugging tool [...] Each line in the trace contains the system call name, followed by its arguments in parentheses and its return value.

A system call is simply what programs use to ask the system to do operations like opening a file, writing in it, creating a now process... etc. It is not necessary to understand every single system call to use strace.

Once the package strace is installed you can either trace a program when launching it, or hook yourself to an already running thread.

Fair warning, strace is quite resource intensive, and will slow down your program further if you simply print its output to your terminal instead of sending it into files, so it's better to send it to a file and "tail -f" it. Also, as strace creates one file per process, so for multi-process app I advise creating a dedicated directory like /tmp/strace_my_program.

Basic commands#

If you want to trace a program/script that as you launch it (arguments are explained below):

1  strace -ffqttv -s1000 -o /tmp/strace_0/strace.log -p <your_command_here>

For instance:

1  strace -ffqttv -s1000 -o /tmp/strace_0/strace.log -p main.py -s "hello"

And if you want to attach yourself to an already running process:

1  strace -ffqttv -s1000 -o /tmp/strace_0/strace.log -p <pid of the process>

The additional arguments I put are:

-ff: follow 'forks' (children processes) and send their output into separate files;
-q: remove strace messages about "attaching" and "detaching", which aren't really useful;
-tt: add timestamps with milliseconds precision to the logs, especially useful to debug multi-process apps;
-v: do not abbreviate calls' details, which can be useful;
-s1000: specify maximum string size (default is 32, which is often too small);
-o /tmp/strace_0/strace.log: send the output to this file: if there are multiple processes, the files will be created as strace.log.<pid>.

I really advised to use these parameters for more useful logs but you can obviously adapt them.

Filtering#

If you know what you're looking for, you can filter the output. For instance, you can filter system calls based on their category:

-e trace=file;
-e trace=process;
-e trace=network;
-e trace=signal;
-e trace=ipc;
-e trace=desc;
-e trace=memory.

And even filter directly the system calls: for instance, see only when the program writes to standard (stdout) and error (stderr) output:

-e write=1,2

You can also narrow down the output based on a file path:

-P /var

Reading the file(s)#

strace is quite verbose, but as its output is standardized, you can use plugins or even settings in your text editor to highlight it. For instance in vim, simply type:

:set filetype=strace

triggering i3lock on sleep

2021-04-22T18:34:00+02:00

adapted from https://wiki.debian.org/ScreenLockingOnSleep

/etc/systemd/system/i3lock.service:

 1  [Unit]
 2  Description=i3lock
 3  Before=sleep.target
 4 
 5  [Service]
 6  User=nemecle
 7  Type=forking
 8  Environment=DISPLAY=:0
 9  ExecStart=/home/nemecle/.bin/scripts/lock.sh
10 
11  [Install]
12  WantedBy=sleep.target

lock.sh:

1  #!/usr/bin/env bash
2 
3  i3lock -i $HOME/Wallpaper_lock.png

nemecl.eu's design

2021-04-21T18:23:00+02:00

The design will be perpetually changing

Content

General idea
- Inspiration board
Typefaces and fonts
- Typefaces
- fonts
Colours
Design

General idea#

Look: 80 hacker zines, terminal look, screen printed zines;

To explore: constructivist look (e.g. El Lissitzky), swiss Punk typography, brutalist architecture.

Execution: monospace typefaces, spaced design, visual hierarchy through size, weight, position and colour rather than decoration, very few colours and generally low contrast colours to imitate print limitations, big blocks of colour.

Inspiration board#

Typefaces and fonts#

Typefaces#

"Inconsolata" almost all elements by now
Formely "Deja Vu Sans Mono Book" for title elements
"Libre Baskerville" as potential non-monospace body typeface

fonts#

Heavy use of various font weights, leveraging the fact that Inconsolata is available from 200 to 900.

Delivery of 900 seems to fail as of now

900: Biggest, for special elements
700: Bold elements (and titles)
400: regular, for text body
200: Ultra thin, for visual elements or very big texts

Example

Billy and his shoes — 20/10/2021 @ THE DIRTY BARN

Colours#

#002a38
text

#213e48
darkest

#294d59
darker

#355a65
dark

#446874
intermediate

#86a6b0
light

#b4c9d0
lighter

#dbe5e9
lightest

#f6f5f0
background

#a54747
accent

Design#

Headings#

Heading 1

Heading 2

Heading 3

Heading 4

Paraphraph: Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Text elements#

drop-cap-word + block:

INTRO

"Frames"#

default version

"full" version

Alert
zis is alert

Aleeeert
aaaaaaaaa

Warning
Things are bad

Waaarning
You may panic

Success
Everything is cool

Succeeeesss
But you're full of it

Blue
Nice blue hey

Bluuuue
Lot of it

Dark
Dark end

So dark
Dark end

Charcoal
Dark end

So Charcoal
Dark end

Details-summary element#

Here is a summary.
this is a summary that can be clicked to view/hide its details

Even though there is not much to see here.

left and right aligned elements#

I'm here here.

And here.

These elements are not meant for this text layout, so they might overlap.

But they are responsive, as they will merge within the text as a float element on tablets, and as a full width block on mobile.

Footnotes are by default right-aligned, but will also turn into a full-width element if the viewport is not wide enough. [1].

[1]	Like so. Likewise, those are not meant for this text layout.

Forms#

Code highlighting#

Here is an inline literal element: cat base_* > base.css

Docutils, the software used by pelican to generate this page from RestructuredText, seems to turn inline literals into <tt> elements which are deprecated instead of <code> elements. Might need to be looked into.

The <pre> blocks are as of now a loose implementation of the Nord colour theme and remained to be defined. The line numbers disappear on hover to avoid interfering with copy-pasting.

Code (without highlighting):

1 usermod nemecle -aG sudo

Code (with highlighting):

 1  #!/usr/bin/python3
 2  """
 3  this is a test script.
 4  """
 5 
 6 
 7  import re
 8  from datetime import datetime
 9 
10  EXTENSION = "PNG"
11 
12  def main():
13      """
14      main loop
15 
16      """
17 
18      print("Hello world")
19 
20      for i in range(0, 10):
21          print(str(i))
22      return
23 
24 
25  if __name__ == '__main__':
26      main()

Setting IPv6 on a server

2021-04-19T18:10:00+02:00

Potentially out of date

Based on this documentation from OVH adapted for my own uses, on a debian-nginx server

First, add to the file /etc/sysctl.conf:

1  net.ipv6.conf.all.autoconf=0
2  net.ipv6.conf.all.accept_ra=0

Then to apply changes:

1 sudo sysctl -p

add to /etc/network/interfaces (or /etc/network/interfaces.d/50-cloud-init.cfg for OVH):

1  iface eth0 inet6 static
2  address YOUR_IPv6
3  netmask 128
4 
5  post-up /sbin/ip -f inet6 route add IPv6_GATEWAY dev eth0
6  post-up /sbin/ip -f inet6 route add default via IPv6_GATEWAY
7  pre-down /sbin/ip -f inet6 route del IPv6_GATEWAY dev eth0
8  pre-down /sbin/ip -f inet6 route del default via IPv6_GATEWAY

For some reasons guides add a indentation but it seems to be counter productive (note to self, read documentation more carefully)

And finally to apply it (make sure to have a KVM on hand in case network interfaces fail to come back):

1 sudo /etc/init.d/networking reload

The reload command gives more verbose logs.

Then add AAAA record on registrar.

And finally add to website configuration (nginx):

1  listen [::]:443 ssl http2 default_server;

And to apply:

1 sudo service nginx reload

Use two different IPs on a single web server

2021-03-22T18:30:00+01:00

see "two ip one vps":

 1 auto ens3
 2 iface ens3 inet dhcp
 3 
 4 auto ens3:0
 5 iface ens3:0 inet static
 6 address <ip#1>
 7 netmask 255.255.255.255
 8 
 9 auto ens3:1
10 iface ens3:1 inet static
11 address <ip#2>
12 netmask 255.255.255.255

Will say "failed to raise interface", but ip addr shows it works fine

https://docs.ovh.com/gb/en/public-cloud/configure_a_failover_ip/

Configure a failover IP with Debian Configuring the network interface

For example purposes, we’re using the following variables:

Network interface = ens3

Failover IP address = your_ip_address

IP alias = 1

First, connect to your instance via the command line and open up your the network configuration file.

# nano /etc/cloud/cloud.cfg.d/99-disable-network-config.cfg

Next, edit the file with the configuration shown below. This will prevent changes from being made to your network configuration automatically.

# network: {config: disabled}

Next, open the network configuration file for editing with the following command:

# nano /etc/network/interfaces.d/50-cloud-init.cfg

Then edit the file with the following configuration:

auto ens3 iface ens3 inet dhcp

auto ens3:0 iface ens3:0 inet static address your_ip_address 0 netmask 255.255.255.255

auto ens3:1 iface ens3:1 inet static address your_ip_address 1 netmask 255.255.255.255

Restarting the network service

Finally, restart the network service with the following command:

service networking restart

Ardour

2021-03-22T18:30:00+01:00

Content

Installation
JACK/QjackCtl
Midi keyboard
plugins
Audio tips
- Recording
- Mixing
- Editing

Installation#

This guide was tested on Debian.

install package:

sudo apt install ardour

follow debian wiki regarding real time capabilities (make sure user is still in the audio group):

Add to the end of the file /etc/security/limits.conf:

@audio                -       memlock         unlimited

Add to file /etc/pam.d/common-session:
```
session required pam_limits.so
```
And you need (contrary to some guides which say that it is a eventuality) to add your own user to the aformentionned audio group:
```
sudo usermod -aG audio nemecle
```

if no sound, https://discourse.ardour.org/t/no-sound/90327/2:

Start QJackCTL and click on setup.

Then locate the box named “Interface” and click the arrow on the right side of it pointing to the right (do not confuse this with the drop down menu arrow that points down). Here you can see the list of sound devices on your system, select the one you want to use and click “Ok”.

Click on “Start” on the QJackCTL interface to start Jack.

Now start Ardour and Select Jack as the audio Backend

In this picture I have selected Presonus Audiobox 1818VSL as the audio device.

Basic plugins

sudo apt install eq10q x42-eq calf-plugins

JACK/QjackCtl#

If you use an audio interface or something, go to settings > advanced and set the input device (might require to fumble around to find the display name for your interface).

The "connect" window acts like a literal connection bay: for instance, select the input device on the left and the virtual amp on the right and connect them. You can then simply build a circuit like:

[guitar -> ] audio interface -> virtual amp -> Ardour -> system output (headphones/speakers)

devices names:

Behringer Q502USB: "hw:1 USB Audio CODEC"
my laptop (output): "hw: PCH,1 (ALC3232)"

Midi keyboard#

https://manual.ardour.org/setting-up-your-system/setting-up-midi/midi-on-linux/

plugins#

Noise gate, reverber, flanger, sidechain compressor...#

sudo apt install calf-plugins

Graphic EQ#

eq10q and x42-eq, both available as packages on debian-based distributions:

sudo apt install eq10q x42-eq

The EQ10Q (under the category "plugins" for some reasons) does the job well.

Synthesiser: Helm#

Geonkick#

(drums) https://github.com/geontime/geonkick install redkite as told then dependencies libjack-jackd2-dev, NOT libjack-dev > apt-get install build-essential cmake qjackctl libjack-jackd2-dev libsndfile-dev rapidjson-dev lv2-dev

-> open issue once installed ?

syn-fusion#

Sound removal#

This program is actually a standalone.

http://isse.sourceforge.net/index.html

Audio tips#

Recording#

44kHz/16bit info https://www.youtube.com/watch?v=cIQ9IXSUzuM A->D->A

More than just avoiding the "red zone", stay well below it: modern 16bit precision means that even a rather "low" sound still has a lot a definition and can be safely amplified. Avoid at all cost a signal that is "too hot" (i.e. close to that danger zone), or dynamics management will be a nightmare.

Guitarix (available as a apt package) does a very good job as a virtual amp, as long as the input device is not garbage (i.e. do not use the default mic input on your computer unless you want to make low-cost Lightning Bolt).

Mixing#

Ensure that individual tracks stay just below -3dB
Compress a bit to kill potential hard sound and keep dynamics under control
EQ:

Boost wide, cut narrow
add high shelf and low shelf to keep only useful frequencies
chase parasites by setting a narrow positive filter, move it around until something starts resonating, and cut it off
boost based on feeling, according to this chart:https://web.archive.org/web/20190417040339/https://www.independentrecording.net/irn/resources/freqchart/main_display.htm

de-essing (reduction of "sss" clashing sounds) on human voice if necessary (plugin to find, check http://www.linuxdsp.co.uk/download/lv2/download_dsr500/index.html and https://twinysam.github.io/FreeAudioPluginList/)
pan & wide

Editing#

The screenshot corresponding to the text above.

The amplitude of the first audio ends at the bottom ,and the next one start at the top: the speaker will quickly transition between the two, creating a pop.

Be careful with "mismatched amplitude" when sticking two audio elements, including when doing video editing. Avoid it by either:

doing "snap-to-zero crossing" (always cutting when the amplitude is at zero, but this is not trivial in a video editing software);
do a small cross fade;
slightly reduce the sound on the last frame of each element;
add in the background the "room tone" (i.e., the background sound of the recorded place without any other sound)

Mediawiki

2021-03-22T18:30:00+01:00

Content

Installation
Update

Installation#

To do during ansible script creation

Update#

First, backup the database. The "set history +o" disables the shell history to avoid storing passwords:

set +o history
sudo mysqldump --user=<user> --password=<password> <database> > /var/backup/<database>/<database>-`date +%d-%m-%y`.sql
sudo mysqldump --user=<user> --password=<password> <database> --xml > /var/backup/<database>/<database>-`date +%d-%m-%y`.xml
set -o history

Then download and extract the release:

cd /tmp
wget https://releases.wikimedia.org/mediawiki/1.35/mediawiki-1.35.1.tar.gz
tar -xvzf mediawiki-1.35.1.tar.gz

Copy the existing data and settings into the new release:

sudo cp -r wiki/LocalSettings.php wiki_1.35/
sudo cp -r wiki/images/ wiki_1.35/
sudo cp -r wiki/extensions/* wiki_1.35/extensions/
sudo cp -r wiki/resources/assets/logo.png wiki_1.35/resources/assets/
sudo cp -r wiki/skins/* wiki_1.35/skins/

Then switch directory

mv wiki wiki_old && mv wiki_1.35 wiki

finally:

cd maintenance/
sudo php update.php

Nextcloud

2021-03-22T18:30:00+01:00

https://websiteforstudents.com/install-nextcloud-on-ubuntu-17-04-17-10-with-nginx-mariadb-and-php/

Content

Basic install
Logging
Troubleshooting
Importing contacts in "Contacts" apps errors
Uploads fail silently on android

Basic install#

Use USB connected drive as data directory#

/etc/fstab:

UUID=360C3D2F0C3CEB8D /var/external_drive ntfs-3g uid=33,gid=1001,rw,user,exec,umask=007 0 0

Then in config/config.php, set "var/external_drive" as directory (might require to install package for NTFS)

NFS#

sudo apt update
sudo apt install nfs-common

test:

sudo mount -t nfs 192.168.1.152:/volume1/data /var/data

fstab :

192.168.1.152:/volume1/data  /var/data   nfs      rw,sync,hard,intr  0     0

nginx#

sudo systemctl stop apache2
sudo systemctl disable apache2

sudo apt install nginx
sudo systemctl stop nginx.service
sudo systemctl start nginx.service
sudo systemctl enable nginx.service

mariadb#

sudo apt install mariadb-server mariadb-client
sudo systemctl stop mariadb.service
sudo systemctl start mariadb.service
sudo systemctl enable mariadb.service
sudo mysql_secure_installation

Enter current password for root (enter for none): Just press the Enter
Set root password? [Y/n]: Y
New password: Enter password
Re-enter new password: Repeat password
Remove anonymous users? [Y/n]: Y
Disallow root login remotely? [Y/n]: Y
Remove test database and access to it? [Y/n]:  Y
Reload privilege tables now? [Y/n]:  Y

sudo systemctl restart mariadb.service

sudo apt install php-fpm php-mbstring php-xmlrpc php-soap php-apcu php-smbclient php-ldap php-redis php-gd php-xml php-intl php-json php-imagick php-mysql php-cli php-ldap php-zip php-curl unzip

php-mcrypt#

DEPRECATED. See this issue

sudo mysql -u root -p
CREATE DATABASE nextcloud;
CREATE USER '<user>'@'localhost' IDENTIFIED BY 'new_password_here';
GRANT ALL ON nextcloud.* TO '<user>'@'localhost' IDENTIFIED BY 'user_password_here' WITH GRANT OPTION;

cd /tmp && wget https://download.nextcloud.com/server/releases/nextcloud-<release>.0.1.zip
unzip nextcloud-17.0.1.zip
sudo mv nextcloud /var/www/html/nextcloud/

sudo chown -R www-data:www-data /var/www/html/nextcloud/
sudo chmod -R 755 /var/www/html/nextcloud/

sudo nano /etc/nginx/sites-available/nextcloud

sudo ln -s /etc/nginx/sites-available/frigg_nemecl /etc/nginx/sites-enabled/

# if network mount:

(select /var/data) (chown /var/data)

Live DNS#

https://github.com/cavebeat/gandi-live-dns

sudo apt update && sudo apt upgrade && sudo apt install unzip python-requests python-args python-simplejson

cd /tmp && git clone https://github.com/cavebeat/gandi-live-dns.git
sudo mv /tmp/gandi-live-dns /opt

# Settings: (https://forum.yunohost.org/t/configuration-nextcloud-php-fpm-pm-max-children/10911) in /etc/php/7.x/fpm/pool.d/www.conf:

pm = static
pm.max_children setting = 8 ; depends on CPU, good on Pi 3

and:

sudo service php7.3-fpm restart

Logging#

Add:

"log_type" => "file",
"logfile" => "nextcloud.log",
"loglevel" => 3,
"logdateformat" => "F d, Y H:i:s",

to config/config.php

https://docs.nextcloud.com/server/latest/admin_manual/configuration_server/logging_configuration.html#file

Troubleshooting#

Cannot login, always goes back to login page#

More often than not, that indicates a permission issue, namely in the nextcloud directory. Simply do (adapt it if needed):

sudo chown -R www-data:www-data /var/www/html/nextcloud/

php settings ignored#

Also applies for any modification like adding Redis (Redis server went away) If some php settings (typically, memory_limit, as shown in settings > Administration > System) seem to be ignored, you might need to restart the php-fpm service as well. For instance:

sudo service php7.3-fpm restart

Error 423#

sudo -u www-data php occ files:scan --all

Error 504 "Error when assembling chunk"#

If some php settings (typically, memory_limit, as shown in settings > Administration > System) seem to be ignored, you might need to restart the php-fpm service as well. For instance:

in /etc/php/7.x/fpm/pool.d/www.conf, if max_execution_time is low (e.g. 30), raise it to 3600 :

max_execution_time = 3600

Then restart fpm service: .. code:

sudo service php7.3-fpm restart

source

nextcloud locked files#

https://help.nextcloud.com/t/file-is-locked-how-to-unlock/1883

You can’t access files because they are locked and you find such errors in your logfile:

{"reqId":"0DijSqEkfOg2iyh9VD8J","remoteAddr":"xx.xx.xx.xx","app":"webdav","message":"Exception: {\"Message\":\"HTTP\\\/1.1 423 \\\"path\\\/file.extension\\\" is locked\",\"Exception\":\"OCA\\\\DAV\\\\Connector\\\\Sabre\\\\Exception\\\\FileLocked\",\"Code\":0,\"Trace\":\"#0

Manually disable locking state:

put Nextcloud in maintenance mode: edit config/config.php and change this line: 'maintenance' => true, Empty table oc_file_locks: Use tools such as phpmyadmin or connect directly to your database and run (the default table prefix is "oc<underscore>", this prefix can be different or even empty): DELETE FROM oc_file_locks WHERE 1 disable maintenance mode (undo first step). Make sure your cron-jobs run properly (you admin page tells you when cron ran the last time): https://docs.nextcloud.org/server/13/admin_manual/configuration_server/background_jobs_configuration.html 1.7k

Permanent solution (if it happens regularly)

on your own server: Use redis for this feature. It is faster and so far no problems have been reported. You can follow the instructions for memory-caching in the docs: https://docs.nextcloud.org/server/13/admin_manual/configuration_server/caching_configuration.html#id4 6.1k Shared hosting (others who can’t install redis): You can disable the file locking, edit your configuration file config/config.php: 'filelocking.enabled' => false, However, disabling is not a good solution. You can run into problems when several processes try to write to a file (especially online editors in the web-interface). In single-user and single-client environments, it’s probably less of a problem.

> sudo service php7.3-fpm restart

In addition to the package nextcloud-desktop, you can search for integrations for file managers. For instance caja-nextcloud for MATE.

Update fails on intergrity and stuck on step 5#

when trying to start update from web interface, updater fails on integrity check, and when page is reloaded it is stuck on "Step 5 is currently in process. Please reload this page later."

Typical root cause#

Check web server logs (for instance, in /var/log/nginx/, do sudo grep -rn "bytes exhausted") if a message "Allowed memory size of xxx bytes exhausted" appears. if so, server ran out of memory when updating and fails to restart updating process.

go into php server setting files (for instance in my case, /etc/php/7.3/fpm/php.ini), and change memory_limit to a higher value, or check php settings ignored

Unlocking updater#

Once the problem is solved, unlock the updater by logging as the owner of the Nextcloud directory (typically www-data), browse to the directory, and launch:

php occ maintenance:repair

The updater should be right back again.

Source: issue #316 on Github

Importing contacts in "Contacts" apps errors#

When trying to import contacts from a .vcf file, all fail (e.g.: 200 contacts = 200 errors).

Probably has to do with vcf version. Follow this forum post: https://help.nextcloud.com/t/import-vcard-2-1-to-address-book/102812

Uploads fail silently on android#

When uploading files from the android file, everything seems fine but files do not appear on server. With Auto-upload, files stop uploading alltogether.

Open side panel, go to "uploads": if files (including auto-uploads) show "Local file not found", this is it.

Open android settings, and go to Apps > Special app access > All files access > Nextcloud (or search "all files access" and enable it.

source: https://www.reddit.com/r/NextCloud/comments/1lu2ba7/local_file_not_found_android_upload/

Siril

2021-03-22T18:30:00+01:00

Content

Bias/offset
Dark
- basic
- most optimal
Lights
Comets
Sources

Bias/offset#

inherent noise of sensor Can be taken once and used for months Temp doesn't matter very much for most cams Take 20-40 frames with lens cap on, shortest exposure available, same ISO

Open Siril, set work dir, add all raws, convert (no debayer /!, keep all channels) Then stacking > median stacking, no normalisation

Dark#

Thermal noise during or After shooting (temp important), same params but with lens cap on

Open Siril, set work dir by clicking on "house" icon"), select sequence name (for instance, "neowise_dark"), add all raws, convert (no debayer /!, keep all channels)

Then go to stacking

basic#

Then stacking > median stacking, no normalisation

most optimal#

average stacking with rejection, no normalisation, Winsorized sigma clipping (if rejection ends up > 0.5%, fall back to median stacking, no norm)

Lights#

Simple photos of the object

In siril, change dir, set seq name, convert WITH debayer

Registration: registration method "Global star", else don't change settings

Comets#

Register sequence using global sking sequence tab, display first image, select comet, then using sequence tab again select last image et pick object in #2

Sources#

basics: https://www.youtube.com/watch?v=qYXd0rrAzjg pre-process: https://www.youtube.com/watch?v=RswLpEE7EZ4 https://free-astro.org/index.php/Siril:Tutorial_preprocessing

Updating linux

2021-03-22T18:30:00+01:00

Content

Updating Linux

Updating Linux#

1 sudo apt upgrade && dist-upgrade && autoremove && autoclean

Web development

2021-03-22T18:30:00+01:00

Content

Writing
Colours
Design
- Typography
- html5
- CSS
Serving
- Default "catch all" nginx server block
- Allow "window.print()" with Content Security Policy
Test
Security
- https/SLL
- CORS
Performance
Accessibility
SEO
- link sharing

Writing#

Colours#

There is no "systematic" approach to colours: it's all a matter of design decision. Do not fall into the whole "triad/split complementaries" trap from palette creation tools. Choose a main "colour" with multiples shades and tints, and a main accent colour. Build a palette from that, and add necessary colours (e.g. warnings and success texts) based on how good they look with the rest.

An interesting article on the subject: Building Your Color Palette on refactoringui.com.

The only framework for palette creation is AA(A) colour constrasts (see Accessibility).

For instance, first choose a text colour and a background color:

#002a38
text

#f6f5f0
background

The choose an "accent" colour for elements that need to pop:

#002a38
text

#f6f5f0
background

#a54747
accent

Then choose intermediate colours, for nuance and more or less important elements:

#002a38
text

#294d59
darker

#446874
intermediate

#b4c9d0
lighter

#f6f5f0
background

#a54747
accent

Depending on your needs, you might require more nuances:

#002a38
text

#213e48
darkest

#294d59
darker

#355a65
dark

#446874
intermediate

#86a6b0
light

#b4c9d0
lighter

#dbe5e9
lightest

#f6f5f0
background

#a54747
accent

And finally add colours for specific cases such as error messages, warning… etc:

#c72323;
red

#eea828;
orange

#e8e852;
yellow

#47b787;
green

#5057ba;
blue

#7d517d;
purple

#ffe3ff;
lila

#88b4a8;
flat-green

#333333;
charcoal

The colours should be defined using CSS's variables to ensure consistency between elements.

Design#

Typography#

Knowledge#

better typography CSS https://betterwebtype.com/articles/2018/10/15/rhythm-in-web-typography/ web typography https://css-tricks.com/typography-for-developers/ webfont guide https://www.freecodecamp.org/news/web-fonts-in-2018-f191a48367e8/ vertical rythm https://www.gridlover.net/try

Notable typefaces#

atypical open-source typefaces http://velvetyne.fr/

Tools#

(Kind of broken) @font-face CSS rules and various format generator (woff woff2 ttf) https://transfonter.org/

Typical workflow#

typeface/font downloader ttf woff woff2 https://google-webfonts-helper.herokuapp.com/fonts/zilla-slab?subsets=latin,latin-ext Add latin-ext if available Copy and paste font-face CSS rules, add font-display: swap;

html5#

favicon#

How to Favicon in 2025: Three files that fit most needs

boilerplate#

Partially inspired by this

 1  <!DOCTYPE html>
 2  <html lang="en" xml:lang="en" xmlns="http://www.w3.org/1999/xhtml">
 3    <head>
 4      <meta charset="UTF-8" />
 5      <meta name="viewport" content="width=device-width,
 6      initial-scale=1.0">
 7      <title>home | nemecl.eu</title>
 8      <base href="https://nemecl.eu/"/>
 9      <meta name="title" content="home | nemecl.eu">
10      <meta name="description" content="Homepage for nemecl.eu. Homemade interwebz, some kind of chaos garden.">
11 
12      <link rel="icon" href="/favicon.ico" sizes="32x32">
13      <link rel="icon" href="/favicon.svg" type="image/svg+xml">
14      <link rel="apple-touch-icon" href="/apple-touch-icon.png"><!-- 180×180 -->
15 
16      <link rel="preload" href="main.css" as="style">
17      <link rel="preload" href="https://rsc.aligot-death.space/css/base.css" as="style"/>
18 
19      <link rel="preload" href="/rsc/webfonts/inconsolata/inconsolata-v20-latin-ext_latin-regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
20 
21      <link rel="stylesheet" href="https://rsc.aligot-death.space/css/base.css" type="text/css" />
22      <link rel="stylesheet" href="main.css" />
23 
24      <link rel="canonical" href="http://nemecl.eu/">
25      <meta property="og:title" content="home | nemecl.eu">    <meta property="og:locale" content="en_GB">
26      <meta property="og:type" content="website">
27      <meta property="og:url" content="http://nemecl.eu/">
28      <meta property="og:description" content="Homepage for nemecl.eu. Homemade interwebz, some kind of chaos garden.">
29      <meta property="og:image" content="https://rsc.aligot-death.space/images/favicon.png">
30 
31      <meta name="twitter:card" content="summary">
32      <meta name="twitter:title" content="nemecl.eu">
33      <meta name="twitter:url" content="https://nemecl.eu">
34      <meta name="twitter:site" content="@nemecle">
35      <meta name="twitter:description" content="Welcome to my small patch of homemade interwebz, some kind of chaos garden.">
36      <meta name="twitter:image" content="https://rsc.aligot-death.space/images/favicon.png">
37      <!--<base href = "https://nemecl.eu"/>-->
38 
39      <meta name="theme-color" content="#eaeaea">
40 
41      <script src="https://rsc.aligot-death.space/scripts/instant-page_5.1.0.js" type="module"></script>
42 
43      <meta name="format-detection" content="telephone=no">
44      <meta name="twitter:dnt" content="on">
45    </head>
46    <body>

<article> is not "article" as in blog post, but as in "shop article": you can use it for instance for a list of articles. An article would simply be in a <main> element, under a <section> in necessary

CSS#

sort CSS properties by alphabetical order https://jerrylowm.medium.com/alphabetize-your-css-properties-for-crying-out-loud-780eb1852153

"classless" CSS is a framework that doesn't use html classes or IDs, and only leverages html elements. It's not necessarly a good idea to apply it to everything but using as little classes and IDs as possible makes writing pages easier.

Avoid margin to simplify layout management: Margin considered harmful

"full-bleed" layout: a flexible layout for width-contrained content, while allowing exceptions. See this article

Serving#

Default "catch all" nginx server block#

from nginx documentation.

 1  http {
 2    index index.html;
 3 
 4    server {
 5      listen 80 default_server;
 6      server_name _; # This is just an invalid value which will never trigger on a real hostname.
 7      access_log logs/default.access.log main;
 8 
 9      server_name_in_redirect off;
10 
11      root  /var/www/default/htdocs;
12    }
13  }

Allow "window.print()" with Content Security Policy#

Need to add sandbox allow-modals to CSP header

Test#

General test for DNS, HTTPS, content...: https://check-your-website.server-daten.de/?q=nemecl.eu

ct.css: check html head informations https://github.com/csswizardry/ct

Security#

https://infosec.mozilla.org/guidelines/web_security https://blog.vnaik.com/posts/web-attacks.html

https/SLL#

certbot cron#

1  0 2 * * 1 certbot certonly --agree-tos -m nemecle@riseup.net -n --webroot -w /var/www/txt.nemecl.eu -d txt.nemecl.eu && service nginx reload

CSP for simple static website#

1  add_header Content-Security-Policy "default-src self https://rsc.aligot-death.space; child-src 'none'; object-src 'none'; require-trusted-types-for 'script';";

CORS#

if CORS "was blocked because of a disallowed MIME type (“text/html”)." for a js file, might 404/40*

Performance#

https://web.dev/measure/

Accessibility#

accessibility test: ctrl+win+enter for windows narrator

https://www.matuzo.at/blog/i-totally-forgot-about-print-style-sheets/ Testing print style: https://www.matuzo.at/blog/i-totally-forgot-about-print-style-sheets/

tota11y, accessiblity tool in the browser

https://webaim.org/resources/contrastchecker/

Simple speech synthesis

SEO#

Test twitter integration: https://cards-dev.twitter.com/validator

Windows

2021-03-22T18:30:00+01:00

Content

Error "80004005" when connecting to a network drive
Oculus Rift & Nvidia drivers
Windows Process audit
Reinstall Windows Store from Powershell

Error "80004005" when connecting to a network drive#

if you can ping (but uses IPv6) and access it through web interface, but Windows Explorer gives "80004005" error, disable IPv6 and flush dns cache by launching from cmd:

1 ipconfig /flushdns

Given that everyone keeps giving ipv6 literal .net, I guess SMB in windows can't ipv6 https://docs.netapp.com/ontap-9/index.jsp?topic=%2Fcom.netapp.doc.cdot-famg-cifs%2FGUID-37002578-913C-4883-8328-731FB1EAB558.html

but might be a protocol mismatch in some cases: try disabling SMB 1.0 (archive)

Oculus Rift & Nvidia drivers#

Oculus "Update your graphics card drivers" : update it from Geforce Experience, not from website DL

Windows Process audit#

https://processhacker.sourceforge.io/

Reinstall Windows Store from Powershell#

https://www.winhelponline.com/blog/restore-windows-store-windows-10-uninstall-with-powershell/

Avoid have to type SSH passphrase everytime

2021-03-22T00:00:00+01:00

from How to avoid being asked passphrase each time I push to Bitbucket

1 ssh-add

"can't read superblock"

2021-03-22T00:00:00+01:00

Not necessarly a drive fault, can for instance mean that it doesn't have enough power for external drive https://askubuntu.com/questions/539880/cant-mount-an-ide-hdd-cant-read-superblock-error-message

Direct kernel communication

2021-03-22T00:00:00+01:00

Can be useful in case of partially stalled or locked system

Alt, press SysRq (The SysRq being the Print Screen key), release SysRq, press <command key>, Hit ALT-SysRq-<command key>, release everything.

Or if possible,

1 echo t > /proc/sysrq-trigger

with t being the command key.

See the wikipedia page for full command key list.

Can't beat a tool that has a "Perform a system crash" feature.

ffmpeg

2021-03-22T00:00:00+01:00

Content

Creating a timelapse
Concatenate ".ts" files from local record

Creating a timelapse#

1080p/30i with jpg files:

1 ffmpeg -framerate 30 -pattern_type glob -i '*.jpg' -c:v libx264 -r 30 -pix_fmt yuv420p -s hd1080 out.mp4

Concatenate ".ts" files from local record#

For instance, if you made a local record using OBS

Create file list:

File order might be wrong, to fix

1 (for %i in (*.ts) do @echo file '%i') > mylist.txt

Then concatenate to a single mp4 file:

1 ffmpeg -f concat -i mylist.txt -c copy output.mp4

Changing the theme for GTK from CLI

2021-03-22T00:00:00+01:00

(for instance, If you use typical GTK utilities in i3)

1 vim $HOME/.config/gtk-3.0/settings.ini

Example config:

1 [Settings]
2 gtk-icon-theme-name = Paper
3 gtk-theme-name = Arc-Darker
4 gtk-font-name = DejaVu Sans 11

Managing multiple screens from CLI/i3wm

2021-03-22T00:00:00+01:00

install xrandr. On debian, do:

sudo apt install x11-xserver-utils

More often than not, this is enough:

xrandr --auto

to see possible settings, type:

xrandr

to add a second screen next to a 2560x1440, plugged on HDMI2:

xrandr --output HDMI-2 --pos 2560x0 --mode 1920x1080 --rate 60

Remove a screen:

xrandr --output HDMI-2 --off

Linux

2021-03-22T00:00:00+01:00

Linux tips and notes.

Linux performance monitoring tools

2021-03-22T00:00:00+01:00

See http://www.brendangregg.com/linuxperf.html

Mounting secondary drive on boot

2021-03-22T00:00:00+01:00

Content

Emergency mode

sudo blkid
sudo umount /mnt/xxx
vim /etc/fstab: if in doubt, just pu UUID, mount point, FS type and "defaults 0 0"

Emergency mode#

Happened to me as I changed my DATA HDD. because I added it to /etc/fstab and the disk wasn't the same anymore, the system would drop into emergency mode.

Once connected, try mounting all mount points and see errors:

mount -a

Easy network management from terminal

2021-03-22T00:00:00+01:00

Might requires to install nmcli on some distros. Make sure to do it before you're disconnected.

the nmcli utility is very easy and useful to manage network settings, including Wi-fi connections.

For instance, listing available Wi-Fi networks:

1 nmcli d wifi list

Connecting to a wifi network:

1 nmcli d wifi connect "Super Wifi" password "John19970612"

pelican

2021-03-22T00:00:00+01:00

Pelican is a static site generator in python which is not necessarily beautiful out of the box, but offers a lot of control on the generation process.

Content

More practical work environment
Writing
Theming
Must have plugins
Troubleshooting
- "TypeError: not all arguments converted during string formatting"

More practical work environment#

Add to ~/.bashrc the following aliases:

1 alias pc="pelican content"
2 alias pl="pelican --listen"

simply build the website py typing pc from the directory, and pl to launch Pelican’s web server.

The website can be build without shutting down the server, so you can simply launch the pl command in another terminal (or typically, in another tmux pane) and build the website on the fly.

Writing#

This part assumes that you use RestructuredText format for your articles.

Easy styling#

to add a class to a paragraph, use (be careful with the whitespaces to avoid errors):

1 .. class:: warning
2 
3 This part assumes that you use RestructuredText format for your articles.

Example: the warning just above (assuming you've created the corresponding CSS class)!

Multiple ..class can be nested, as long as the empty lines are respected. For instance:

1 .. class:: align-right
2 
3 .. class:: side-note
4 
5 **Le Règlement Général sur la Protection des Données (RGPD) c'est quoi ?**

using html within a .rst file#

Use:

1 .. raw:: html
2 
3    <details>
4    <summary><b>Here is a summary.</b><br>
5    this is a summary that can be clicked to view/hide its details </summary>
6    Even though there is not much to see here.
7    </details>

Result:

Here is a summary.
this is a summary that can be clicked to view/hide its details

Even though there is not much to see here.

Table of content#

Simply add above your first part title (or anywhere, really):

1 .. contents:: <ToC title>
2 :depth: 2
3 :backlinks: none

Result:

Content

More practical work environment
Writing
Theming
Must have plugins
Troubleshooting
- "TypeError: not all arguments converted during string formatting"

Linting code#

You can find the list of available languages for the code block here Configuration =============

Link management#

To ensure proper links, working page anchors and website structure:

Add <base href="{{ SITEURL }}/{{ output_file }}"/> to HTML header
add RELATIVE_URLS = False to pelicanconf.py (and do not override it to True in publishconf.py)
Make sure to use {{ SITEURL }}/ in templates, such as href = "{{ SITEURL }}/{{ article.url }}"
Use relative links prefixed with "/" to ensure that they start from the root of your website.

clean URLs#

In the pelicanconf.py file, set the following variables:

TAGS_URL = "tags"
CATEGORIES_URL = "categories"
ARCHIVES_URL = "archives"

ARTICLE_URL = "{slug}"
ARTICLE_LANG_URL = '{slug}-{lang}'

PAGE_URL = "{slug}"
PAGE_SAVE_AS = "{slug}.html"
PAGE_LANG_URL = 'pages/{slug}-{lang}'

CATEGORY_SAVE_AS = 'categories/{slug}.html'
CATEGORY_URL = 'categories/{slug}'
TAG_SAVE_AS = 'tags/{slug}.html'
TAG_URL = 'tags/{slug}'

The result will be urls of the form example.com/making-soup or example.com/categories/linux.

To ensure that these links are used, you also need to configure the web server. In the case of nginx:

location = / {
  index index.html;
}

location / {
    root /var/www/example.com/;
    try_files $uri.html $uri =404;
}

Theming#

Using article "summary"#

Use the striptags filter to avoid having <p class = "first"> garbage in your meta description:

1 {% if  article.summary  %}
2 <meta property="og:description" content="{{ article.summary|striptags|safe  }}">
3 {% endif %}

Article ordering#

By default articles are ordered in "reverse-date". But any metadata (with a "reverse-" prefix to... reverse) can be used: for instance, this wiki uses "title", like this in pelicanconf.py:

1 ARTICLE_ORDER_BY = "title"

See Ordering content for more details.

Translation switch for custom pages#

To get translation links for custom page (:save_as: and :template:), edit the translations.html from the theme and the following new macro:

 1 {% macro translations_for_page(page) %}
 2 <div>
 3 {% if page.translations %}
 4 <span class = "block">{{ page.lang }}</span>
 5 {% for translation in page.translations %}
 6 <a class = "inverted-block" href="{{ SITEURL }}/{{ translation.url }}" hreflang="{{ translation.lang }}">{{ translation.lang }}</a>
 7 {% endfor %}
 8 {% endif %}
 9 </div>
10 {% endmacro %}

This new macro is called translation for page. And then call it using the following jinja code (or copy-paste it from another template and adapt it):

1 {% import 'translations.html' as translations with context %}
2 {% if translations.translations_for_page(page) %}
3   {{ translations.translations_for_page(page) }}
4 {% endif %}

For good mesure, here is the content of the file content/pages/index_wiki-fr.rst, which corresponds to the french homepage of this wiki:

 1 Wiki
 2 ########
 3 
 4 :date: 2021-09-29
 5 :template: custom/index_wiki
 6 :save_as: wiki/index-fr.html
 7 :slug: wiki/index
 8 :category: wiki
 9 :lang: fr
10 :status: published

Custom home page#

To write properly

Do not modify the "index.html" file hoping to change the homepage: this file is also used for the category, tags and archives pages, which will probably be messed up if you change it. Instead, create a custom template.

follow https://stackoverflow.com/questions/55363180/how-do-i-choose-a-category-page-to-be-the-home-page-for-a-pelican-site /!template "homepage" must be in content/templates and exclude it, as weird as it is then, https://stackoverflow.com/questions/19283880/querying-for-specific-articles-via-tag-category-in-pelican-themes

Navigation menu with article list#

<-- As an example, this wiki uses this.

Both the default (and most popular) theme "notmyidea" and the basic boiletplate theme "simple" gives a confusing example: the way they iterate on categories is:

1 {% for cat, null in categories %}
2 <li{% if cat == category %} class="active"{% endif %}><a href="{{ SITEURL }}/{{ cat.url }}">{{ cat }}</a></li>
3 {% endfor %}

It would be tempting to add a loop under it iterating every articles and checking if the current category being processed is the same, but as it turns out the return value being negated by the "null" is the list of the articles in the given category. So simply do:

1 {% for cat, category_articles in categories %}
2   <li><a class = "category" href="{{ SITEURL }}/{{ cat.url }}">{{ cat }}</a>
3     <ul>
4         {% for artic in category_articles %}<li><a href = "{{ artic.url }}">{{ artic.title }}</a></li>
5         {% endfor %}
6     </ul></li>
7 {% endfor %}

Articles listed per year#

A common way to list articles on personal websites nowaday is to list them by year like so:

2021

I saw a nice rock yesterday

2020

I'm giving up on that blog
More articles to come this year!

2019

A long essay on consciousness
New blog!

To do so in pelican, here is a skeleton template to use within templates.index.html:

 1 <ul class = "article-list">
 2 {% for year, year_posts in articles_page.object_list|groupby('date.year')|reverse %}
 3   <h2>{{year}}</h2>
 4       {% for article in year_posts %}
 5           <li>
 6             <article>
 7               <header>
 8                 <h2>
 9                   <a href="{{ SITEURL }}/{{ article.url }}" rel="bookmark" title="Permalink to {{ article.title|striptags }}">
10                     {{ article.title }}
11                   </a>
12                 </h2>
13               </header>
14             </article>
15           </li>
16       {% endfor %}
17   {% endfor %}
18 </ul>

Explanation: the heavylifting is done by the articles_page.object_list|groupby('date.year')|reverse which groups articles per year with the newest ones at the top. The resulting list is unpacked as the year + the list of articles for that year, the latter being iterated on by a second loop.

List x articles#

If for instance you only want to list the first three articles of a category, you can use the counter tag like so (in this example, 3 articles will be listed):

1   <div class = "image-gallery">
2     {% set counter = namespace(value=0) %}{% for article in articles if ("miscellaneous_digital_drawings" in article.subcategories and counter.value < 3) %}<a href = "{{ SITEURL }}/{{ article.url }}"><img src="{{ article.cover_image }}" alt="{{ counter.value }}"></a>{% set counter.value = counter.value + 1 %}
3     {% endfor %}
4   </div>

Must have plugins#

Refer yourself to the `pelican-plugins<https://github.com/getpelican/pelican-plugins/>`__ repository for more informations

headerid: adds relative links to section headers

Troubleshooting#

"TypeError: not all arguments converted during string formatting"#

https://stackoverflow.com/questions/54047848/pelican-not-all-arguments-converted-during-string-formatting/55130515#55130515

Post-install

2021-03-22T00:00:00+01:00

1 usermod nemecle -aG sudo

To install missing firmware (as printed during boot sequence), this command might be enough:

1  sudo apt install firmware

Basic packages to install:

1 sudo apt install vim sysstat mlocate tmux python3-pip htop rofi xbindkeys keepassxc rsync fd-find hwinfo gparted curl

Recovering from deleted /boot/efi (dual boot)

2021-03-22T00:00:00+01:00

the "/boot/efi" partition used to boot is not exclusive to Linux: it is shared with Windows, and thus you can (and should) provide it to utility tools that fix boot. It is usually on /dev/sda2. You can ls -l /boot/efi from a functioning linux to check its content.

The command to "install" grub requires to provide the disk, not a specific partition (e.g. /dev/sda, not /dev/sda2)

The following lines are a crude hint, the full operation might require some other mounting.

First, create a live USB/CD system, and boot on it.

Then we mount the dead system to be able to work with it:

1 for i in /dev /dev/pts /proc /sys /sys/firmware/efi/efivars /run; do sudo mount -B /mnt$i; done

And we chroot to act as if we were on the dead system:

1 chroot /mnt

Then we mount the EFI partition:

1 mount /dev/sda2 /boot/efi

And finally we execute the "grub-install" command: The bootloader-id argument will be the name appearing in the EFI boot list if you go to the boot menu. Type the following command very carefully, a lot of people online seem to have errors simply due to typos:

1 sudo grub-install --target=x86_64-efi --efi-directory=/boot/efi --bootloader-id=debian --recheck --debug /dev/sda

You may eventually do (once you left chroot with exit like any shell instance):

1 update-grub

Safe bash script boilerplate

2021-03-22T00:00:00+01:00

from full boilerplate here https://betterdev.blog/minimal-safe-bash-script-template/

Add right after shebang:

1 set -Eeuo pipefail

Using SSH as a VPN

2021-03-22T00:00:00+01:00

1 ssh -NTCD 12345 SSH_remote_host_IP

Then in firefox Preference > Network Setting > Settings… proxy as socks with 127.1:12345 Or for a SSH

1 ssh -o ProxyCommand='nc -x localhost:12345 %h %p' username@Far_Away_Host

from: https://blog.gwlab.page/vpn-over-ssh-the-socks-proxy-8a8d7bdc7028

or sshuttle

SSH tunnel

2021-03-22T00:00:00+01:00

ssh -p 34423 -f nemecle@frigg.nemecl.eu -L 5001:192.168.1.152:5000 -N

Then open browser and go to http(s)://localhost:5001

Use "thin non-breaking space"

2021-03-22T00:00:00+01:00

Useful for french typography.

On Plasma Desktop:

Settings > Keyboard > Key Bindings
check "Configure keyboard options"
search for "non-breaking"
check "non-breaking space at the third level for Alt-Gr + Space

(for other combinations like Shift + Space, see https://forums.linuxmint.com/viewtopic.php?t=362965)

vim battle hardened config

2021-03-22T00:00:00+01:00

" BASIC CONFIG
set encoding=utf-8
syntax on
set nu
set rnu
set hls
set sm
set smarttab
set nocompatible
set scrolloff=30
set backspace=indent,eol,start " backspace over everything in insert mode
set ignorecase  " do case insensitive search
set incsearch   " show incremental search results as you type


" STYLE
" colorscheme peachpuff
highlight LineNr ctermfg=black ctermbg=grey
let g:indentLine_color_term = 239


" MAPPING
noremap <space> :
nnoremap <F2> za " Enable folding with F2

" Always show statusline
set laststatus=2
set t_Co=256"


set foldmethod=indent
set foldlevel=99

set tabstop=4
set softtabstop=4
set shiftwidth=4
set textwidth=79
set wrapmargin=0
set expandtab
set autoindent
set fileformat=unix

" web
au BufNewFile,BufRead *.html,*.css,*js
    \ set tabstop=2 |
    \ set softtabstop=2 |
    \ set shiftwidth=2

autocmd BufRead,BufNewFile /etc/nginx/sites-*/* setfiletype conf

" to test
" Bind  key to run python3
" map  :w\|!python3 %
" " Bind  key to run doctests in a python3 module
" map  :w\|!python3 -m doctest %
" " Bind  key to run doctests with verbose output
" map  :w\|!python3 -m doctest -v %
" " Bind  key to run pep8 Python style checker
" map  :w\|!pep8 %

WAcom tablet configuration

2021-03-22T00:00:00+01:00

CLI commands for left-handed, second screen excluded:

1 xsetwacom set "Wacom Bamboo 16FG 4x5 Pen stylus" rotate HALF
2 xsetwacom set "Wacom Bamboo 16FG 4x5 Pen eraser" rotate HALF
3 xsetwacom set "Wacom Bamboo 16FG 4x5 Pen stylus" MapToOutPut 1920x1080+0+0
4 xsetwacom set "Wacom Bamboo 16FG 4x5 Pen eraser" MapToOutPut 1920x1080+0+0

To debug: (from Measuring level of pressure in wacom, tablets or other devices)

Windows: java JWinPointer.
Linux sudo apt-get install evemu-tools evtest, then sudo evtest (sudo is necessary)

the GIMP configuration: https://digimend.github.io/support/howto/apps/gimp/

Simple log analyzer for web traffic

2021-03-22T00:00:00+01:00

sudo apt install goaccess

Then use like:

cd /var/log/nginx
goaccess example.com.access.log

If your logs are setup as "combined" (multiple vhosts per file):

goaccess example.com.access.log --log-format=COMBINED

To concatenate multiple files:

sudo zcat aligot-death.space.access.log.*.gz | goaccess --log-format=COMBINED