من به شخصه آینده روشنی در پس این جنگ برای ایران نمی‌بینم. نه چشم‌انداز روشنی برای اصلاح رفتار و ساختار حکومت می‌بینم، نه افقی روشن برای یک تغییر اساسی در نظام حکمرانی کشور. به همین دلیل، خیلی صریح و بدون واهمه از قضاوت یا کوبیده شدن توسط موافقان جنگ می‌گویم که من با مداخله نظامی به این شکل موافق نبودم. اما حالا که این اتفاق رخ داده، پرسش اصلی برای من دیگر این نیست که چه موضعی درست‌تر بوده، بلکه این است که در دل این جنگ و در فردای آن، چه باید کرد.

برای من، مهم‌ترین مساله در این لحظه نه اثبات حقانیت یک موضع، بلکه تلاش برای فهمیدن پیچیدگی موقعیتی است که در آن قرار گرفته‌ایم.

برای خیلی‌ها دفاع از مداخله نظامی به این معنا نیست که جنگ هیچ هزینه‌ای برای مردم ندارد و حتی به این معنا نیست که حتما در صورت مداخله اوضاع بهتر می‌شود. بسیاری از کسانی که موافق این مداخله هستند، کاملا می‌پذیرند که ممکن است اوضاع حتی بدتر هم بشود. بحث بر سر این نیست که «هر چه بشود بهتر است»، بلکه دقیقا بر سر انتخاب میان چند سناریوی بد است.

یک سناریو، تداوم وضعیت موجود است؛ با هزینه‌های انباشته سرکوب، فرسایش اقتصادی، مهاجرت گسترده، انسداد سیاسی و اجتماعی، و روندی که در بلندمدت می‌تواند به فروپاشی‌های کنترل‌نشده منجر شود.

سناریوی دیگر، مداخله نظامی است؛ با همه ریسک‌های جدی آن: بی‌ثباتی، خشونت، تلفات انسانی، و حتی احتمال تکرار تجربه‌هایی شبیه عراق، سوریه یا لیبی.

استدلال موافقان مداخله این نیست که این ریسک‌ها وجود ندارد، بلکه این است که هزینه‌های سناریوی اول برای آن‌ها قطعی، مستمر و رو به افزایش است، در حالی که سناریوی دوم، با وجود ریسک بالا، یک امکان گسست و تغییر واقعی را باز می‌کند.

در واقع، برای این گروه، مساله انتخاب میان «زوال تدریجی تقریبا قطعی» و «ریسک یک جهش پرهزینه با احتمال تغییر» است.

ممکن است کسی این قمار را نپذیرد، و این موضع کاملا قابل احترام و قابل دفاع است. اما اگر کسی آن را بپذیرد، موضعش لزوما از جنس توهم یا بی‌توجهی به رنج مردم نیست؛ بلکه از جنس ترجیح دادن یک ریسک بزرگ به یک بن‌بست فرساینده است.

در عین حال، به همان اندازه باید گفت که نظر مخالفان جنگ هم کاملا قابل درک است.

مخالفت با جنگ به معنای دفاع از وضعیت موجود نیست. بسیاری از مخالفان، بیش از هر چیز نگران ویرانی اجتماعی، فروپاشی نهادها، گسترش خشونت و رنجی هستند که ممکن است نسل‌ها ادامه پیدا کند. تجربه تاریخی منطقه این ترس را برایشان کاملا واقعی کرده است.

اما شاید تصویر جامعه ایران را هم نتوان تنها در قالب این دو نگرانی توضیح داد. در کنار کسانی که تغییرات بنیادین را، با همه ریسک‌هایش، تنها راه گریز از فرسایش می‌بینند، و در کنار کسانی که از فروپاشی و بی‌ثباتی هراس دارند، بخش قابل توجهی از جامعه نیز همچنان به جمهوری اسلامی و ایده سیاسی و تاریخی پشت آن باور دارد. برای این گروه، جمهوری اسلامی صرفا یک حکومت مستقر نیست، بلکه بخشی از هویت سیاسی، دینی یا تاریخی آن‌هاست؛ چیزی که بقایش را با استقلال، ثبات یا حتی معنای ایران معاصر گره‌خورده می‌بینند. به همین دلیل، افق آینده را نه در فروپاشی کامل این نظم، بلکه در تداوم، بازسازی یا اصلاح آن جست‌وجو می‌کنند.

به‌نظرم بخش مهمی از تفاوت این دیدگاه‌ها، نه از ناآگاهی یا بی‌اخلاقی یک طرف، بلکه از تفاوت در تجربه زیسته و زاویه نگاه آن‌ها می‌آید.

کسی که سال‌ها فرسایش تدریجی زندگی، سرکوب، ناامیدی و انسداد را به‌عنوان خطر اصلی تجربه کرده، ممکن است ریسک یک گسست بزرگ را قابل تحمل‌تر ببیند.

در مقابل، کسی که فروپاشی اجتماعی، جنگ داخلی، آوارگی و ویرانی منطقه را خطر نزدیک‌تر می‌بیند، طبیعی است که با هر نوع مداخله نظامی مخالفت کند.

و در سوی دیگر، کسانی هم هستند که با وجود همه بحران‌ها، ناکارآمدی‌ها و نارضایتی‌ها، همچنان از جمهوری اسلامی دفاع می‌کنند و حاضرند برای حفظ آن هزینه بدهند؛ نه لزوما چون همه وضعیت موجود را مطلوب می‌دانند، بلکه چون فروپاشی این نظم را مساوی با فروپاشی نوعی هویت سیاسی، تاریخی یا حتی امنیت جمعی می‌بینند. نادیده گرفتن این بخش از جامعه، حتی اگر کسی با آن همدل نباشد، فهم ما از نیروهای واقعی و مسیرهای ممکن آینده ایران را ناقص می‌کند.

این اختلاف، بیش از آنکه نزاعی ساده میان درست و غلط باشد، اختلاف بر سر این است که هر گروه کدام نوع خطر را واقعی‌تر، فوری‌تر و تحمل‌ناپذیرتر می‌بیند.

اشکال اصلی جایی شروع می‌شود که هر دو طرف، موضع طرف مقابل را به یک کاریکاتور تقلیل می‌دهند.

اینکه نظر موافقان مداخله نظامی به «بدتر از این نمی‌شود» یا «جنگ هیچ هزینه‌ای ندارد» فروکاسته شود، همان‌قدر نادرست است که نظر مخالفان جنگ به «دفاع از جمهوری اسلامی» یا «رضایت از وضعیت موجود» تقلیل داده شود.

وقتی پیچیدگی و گوناگونی آدم‌های موافق و مخالف حذف می‌شود، نقد کردن موضع آن‌ها راحت‌تر می‌شود، اما دیگر نقدِ نسخه واقعی آن موضع نیست؛ نقد یک تصویر ساده‌شده و ناقص است.

شاید یکی از مهم‌ترین کارهایی که باید همین امروز، در دل این بحران و در اوج شکاف‌ها و خشم‌ها انجام شود، ساختن یک گفتار مشترک برای باهم‌زیستن باشد.

گفتاری که از ضرورت تاریخی برپایی دوباره ایران بر محور ایده باهم‌زیستن دفاع کند؛ اینکه با وجود همه اختلاف‌های عمیق، خشم‌ها، زخم‌ها و تجربه‌های متفاوت، آینده‌ای برای این سرزمین بدون پذیرش تکثر دیدگاه‌ها و بدون امکان زندگی در کنار یکدیگر ساخته نمی‌شود.

اگر قرار است از این وضعیت عبور کنیم، این عبور فقط سیاسی نیست؛ باید دوباره یاد بگیریم که همدیگر را بفهمیم، پیش از آنکه یکدیگر را حذف کنیم.

شاید مهم‌ترین وظیفه امروز ما همین باشد: حفظ امکان باهم‌ماندن، حتی در اوج اختلاف.

چون ایرانِ پس از جنگ، هر شکلی که به خود بگیرد، ناگزیر خانه مشترک همه ما خواهد بود؛ و اگر زبان گفت‌وگو و امکان باهم‌زیستن را از همین امروز نسازیم، هیچ تغییر سیاسی به‌تنهایی آینده‌ای روشن برای این سرزمین نخواهد ساخت.

آینده ایران نه فقط در میدان سیاست، که در توان ما برای باهم‌ماندن، باهم‌شنیدن و باهم‌ساختن رقم می‌خورد.

Here I want to explain why, after a brief and limited restoration of connectivity, messages sent from inside Iran are more easily received outside the country, while the reverse path — messages sent from outside Iran — typically have more difficulty reaching their destination.

Asymmetric Routing

During a crisis or period of restrictions, internet routes typically become asymmetric. This means the outbound path from Iran (sending data to the outside) is less restricted, while the inbound path to Iran (receiving data from outside) is more heavily filtered or blocked. The reason is that controlling incoming information is more important to surveillance systems.

Simply put, “sending” becomes easier than “receiving.” The result of this asymmetry is that a request leaves from inside the country, but the response from outside may never return.

Internet routes become asymmetric (Asymmetric Routing) during outages. The outbound path from Iran is usually more open while the inbound path is more restricted. On the other hand, VPNs and proxies have a harder time re-establishing inbound connections on an unstable internet, especially if they use UDP or have weak keep-alive mechanisms. Also, DNS and external servers sometimes send responses via a route whose return path to Iran is blocked.

Stateless Firewall

A stateless firewall checks each internet packet completely independently. It doesn’t care what was exchanged before this packet. It only looks at whether this packet matches the defined rules.

Advantages

Simplicity is the most important advantage of this type of firewall. It is easy to implement, consumes few resources, and has high processing speed. For this reason, it is suitable for simple routers or networks with very high traffic volumes.

Disadvantages

The main problem with this firewall is that it has no “context.” It cannot understand what connection a packet is part of. For this reason, its rules must either be very permissive (which reduces security) or very strict (which causes legitimate connections to be dropped), and it is not well-suited for precise control of inbound traffic.

Stateful Firewall

A stateful firewall, unlike a stateless one, has memory. When a connection is initiated from inside the network, the firewall records it in a table called the state table. After that, only packets that are a continuation of that recorded connection are allowed through.

If a packet arrives from outside for which no prior connection exists, the firewall treats it as suspicious and typically blocks it.

Advantages

Higher security is the main advantage of a stateful firewall, because it only allows the continuation of connections that started “legitimately.” Control of inbound traffic is much more precise and the risk of abuse is reduced.

Disadvantages

This type of firewall is more complex and consumes more resources. Additionally, it is sensitive to network instability. If a connection is interrupted and reconnects, or a packet is lost, the state may be cleared and subsequent packets may be incorrectly blocked.

Summary Comparison of Stateless and Stateful

In short, stateless is simple and fast but has no awareness of connections. Stateful is smarter and more secure but can drop legitimate connections on an unstable internet. Iran’s internet restrictions more likely use stateful firewalls, because controlling inbound traffic is their priority.

In Iran’s internet restrictions, stateful firewalls are probably used more — meaning if a connection is initiated from inside the country, the response from outside comes back more easily, but if someone tries to send a message from outside without a prior connection from inside, the firewall typically blocks it.

That is why messages sent from inside to outside have a better chance of arriving than messages sent from outside to inside.

TCP and Its Key Messages

Most internet communications (web, messaging apps, email) run on a protocol called TCP. Before sending data, TCP has a specific process for establishing a connection.

First, a packet called SYN is sent, meaning “I want to establish a connection.” The other party responds with SYN-ACK, meaning “I received your request and I’m ready.” The sender then confirms with ACK and the connection is established. After that, the actual DATA is exchanged. Finally, one of the two parties announces with FIN that the connection is over and it is closed.

This defined structure allows stateful firewalls to determine which connections are legitimate.

UDP and the Concept of Pseudo-State

UDP, unlike TCP, has none of these steps. It has no defined start, no acknowledgment, no end. Each UDP packet is sent independently, without the other party necessarily sending a response.

To avoid completely blocking UDP, stateful firewalls create something called a pseudo-state. This means if a UDP packet is sent from inside the network to a specific destination, the firewall temporarily allows a response from that same destination to return. If this time window expires or the internet is unstable, subsequent packets arriving from outside are no longer recognized as legitimate and are dropped.

For this reason, UDP-based connections — such as some VPNs, voice and video calls — become unstable very quickly when internet connectivity is intermittent.

Summary

The fact that messages from inside Iran reach outside more easily but not vice versa is the result of a combination of factors: the asymmetric nature of internet routes, the widespread use of stateful firewalls, the sensitivity of these firewalls to network instability, and the inherent differences between TCP and UDP. This behavior is neither a bug nor accidental — it is the product of design and policy decisions at the network infrastructure level.

One of the greatest advantages of GNU/Linux is its customizability. Many users personalize their system by modifying configuration files. As we use our system over time, these files change continuously, and when their number grows and you use them across different systems, managing them becomes more difficult. One of the best methods for managing these files is using Git.

Window Managers and Their Greatest Advantage: Customizability

I have been using GNU/Linux on my personal systems for a long time and have preferred minimal window managers over large desktop environments like GNOME and KDE for years. Using window managers has made my system lighter and fully customized for my use.

When you use window managers instead of large desktops, you have to configure many system settings and tools yourself.

This makes you better understand the system you use and allows you to change it according to your needs.

Managing configuration files for use across different systems with different hardware and operating systems may seem difficult. Using a Git bare repository is one of the best and most elegant solutions for this. Since most of us already use Git to manage our projects, this approach requires no new tool installation or learning.

Other Tools

Many tools exist for managing dotfiles. Perhaps the most well-known is GNU Stow. A list of these tools can be found here or on the Arch Wiki.

The Git bare repo Method

The first place I found this very clever method was in this Hacker News discussion. After that I read about it in various places and decided to use this method for managing my dotfiles.

Advantages

• No need to install extra tools
• No symlinks required
• Files are managed in a Git repository
• You can use different branches for different systems
• You can easily replicate your configuration when setting up a new system

How Does It Work?

In this method, the Git bare repository is stored in a side folder (such as $HOME/.cfg or $HOME/.myconfig). Using a special alias, Git commands are run against this side folder and won’t interfere with dotfiles or other Git repositories.

Starting from Scratch

If you haven’t previously tracked your configurations in a Git repository, you can easily start using this method with these commands:

git init --bare $HOME/.cfg
alias config='/usr/bin/git --git-dir=$HOME/.cfg/ --work-tree=$HOME'
config config --local status.showUntrackedFiles no
echo "alias config='/usr/bin/git --git-dir=$HOME/.cfg/ --work-tree=$HOME'" >> $HOME/.config/shell/.aliasrc

The first command creates a folder at ~/.cfg which is a Git bare repository that tracks our files.

Then we create an alias so that whenever we want to interact with the dotfiles repository, we use it instead of regular git.

We set a local flag to hide files we are not yet tracking. This way, when you type config status and other commands, files you don’t want to track won’t show up as untracked. I prefer to use echo "*" > ~/.gitignore instead. With this approach, you must use the -f flag to add files in config commands, e.g.: config add -f .dotfile. This helps prevent accidentally adding and committing extra files with config add ..

I use zsh and have sourced the ~/.config/shell/aliasrc file in .zshrc. Note that if you use a different shell, add the alias for that shell in the fourth line:

# for bash
echo "alias config='/usr/bin/git --git-dir=$HOME/.cfg/ --work-tree=$HOME'" >> $HOME/.bashrc

# for zsh
echo "alias config='/usr/bin/git --git-dir=$HOME/.cfg/ --work-tree=$HOME'" >> $HOME/.config/zsh/.zshrc

For fish, you can use this function:

function config -w git -d "Manages dotfiles"
    git --git-dir=$HOME/.dot --work-tree=$HOME $argv
end

Installing Files on a New System or Migrating Existing Files

If you have previously stored your configurations in a Git repository, on a new system you can migrate to this method with the following steps:

Before installing, make sure you’ve added the alias to .bashrc or .zshrc in your repository.

echo "alias config='/usr/bin/git --git-dir=$HOME/.cfg/ --work-tree=$HOME'" >> $HOME/.config/shell/.aliasrc

And that your source repository ignores the folder you’re using as git bare to avoid strange problems: echo ".cfg" >> ~/.gitignore or as mentioned above: echo "*" >> ~/.gitignore

Now clone your configuration repository as a bare repository into a hidden folder at $HOME/.cfg:

git clone --bare <git-repo-url> $HOME/.cfg

Define the alias in your current shell so you can use it:

alias config='/usr/bin/git --git-dir=$HOME/.cfg/ --work-tree=$HOME'

Check out the actual files using config checkout. I use the home branch in my repository to store these files, so for me the command would be: config checkout home. If you use different branches for different systems, make sure you check out the right files.

If there are already similar configuration files in your home directory, the above command will fail with an error. You’ll need to delete or move those files. You can list all files in the repository with this command:

config ls-tree --full-tree --name-only -r <YOUR BRANCH>

#for me
config ls-tree --full-tree --name-only -r home

Then configure this setting to hide untracked files: config config --local status.showUntrackedFiles no

Congratulations! From now on you can manage your files like this:

config status
config add .vimrc
config commit -m "Add vimrc"
config add .bashrc
config commit -m "Add bashrc"
config push

If you used * in .gitignore instead of the status.showUntrackedFiles setting, the commands will look like this:

config status
config add -f .vimrc
config commit -m "Add vimrc"
config add -f .bashrc
config commit -m "Add bashrc"
config push

Resources:

• Hacker News
• Atlassian’s Blog
• Coffee Addict

My Dotfiles

Here I try to write about keyboard configuration in the tty console environment and graphical environments X11 and the Sway window manager on Wayland. Keyboard configuration on Wayland depends on the compositor you’re using, but most of what is discussed here can also be applied on other Wayland compositors, especially those based on wlroots.

Different Keyboards and Layouts

Most of us use the English and Persian QWERTY layout on our keyboards, an example of which can be seen in the image below.

But there are other layouts for QWERTY keyboards depending on the language. For example, the French AZERTY keyboard, which makes changes to the English QWERTY layout for better compatibility with the French language:

For more information, you can read the relevant page on Wikipedia.

There are also other English keyboard layouts, of which DVORAK and COLEMAK are the most well-known.

Example of the DVORAK keyboard:

Example of the COLEMAK keyboard:

The standard Persian keyboard is also classified as a QWERTY keyboard, with the following key layout:

Keyboard Configuration in tty

You may rarely need to change keyboard settings in the console (tty) environment. Most of us use the default English QWERTY layout. However, as mentioned, other layouts exist for English as well. This is one reason you might want to use a different layout in the console.

If you use Systemd as your init system, you can see a list of available layouts with localectl list-keymaps. For me, this command shows 232 different keyboards.

$ localectl list-keymaps | wc -l
232

You can also search for a specific keyboard using grep:

If you don’t use Systemd, the find command can give you a list of available keyboards. The output for me looks like this:

$ find /usr/share/kbd/keymaps/ -type f -name "*"
/usr/share/kbd/keymaps/amiga/amiga-de.map.gz
/usr/share/kbd/keymaps/amiga/amiga-us.map.gz
...

To find a specific keyboard, replace * with the search term surrounded by * characters:

$ find /usr/share/kbd/keymaps/ -type f -name "*uk*"
/usr/share/kbd/keymaps/atari/atari-uk-falcon.map.gz
/usr/share/kbd/keymaps/i386/dvorak/dvorak-uk.map.gz
/usr/share/kbd/keymaps/i386/dvorak/dvorak-ukp.map.gz
/usr/share/kbd/keymaps/i386/qwerty/uk.map.gz
/usr/share/kbd/keymaps/mac/all/mac-uk.map.gz
/usr/share/kbd/keymaps/sun/sunt5-uk.map.gz
/usr/share/kbd/keymaps/sun/sunt6-uk.map.gz

To set your desired keyboard, use the loadkeys command. Note that running this command requires root access:

$ sudo loadkeys fr
$ sudo loadkeys i386/azerty/fr.map.gz

You can use either the keyboard name or the full path to the keymap file. The keyboard you choose must be suitable for the console and its characters must be supported. If you try to set the Persian keyboard for the console, you will encounter an error:

$ sudo loadkeys fa
unicode keysym out of range: U+FDFC
syntax error, unexpected ERROR, expecting NUMBER or LITERAL or PLUS or UNUMBER

To make these settings permanent, you can use the KEYMAP keyword in the /etc/vconsole.conf file.

$ cat /etc/vconsole.conf
FONT=ter-224b
KEYMAP=us

Keyboard Configuration in X11

setxkbmap

For configuring the keyboard in graphical environments that use X11 server, each environment like GNOME or KDE has its own dedicated program. A more general and simpler way is to use the setxkbmap command.

I use this command as follows:

$ setxkbmap -model thinkpad us,ir -option "grp:shifts_toggle,caps:escape_shifted_capslock,altwin:prtsc_rwin,lv3:ralt_switch"

Keyboard configuration with setxkbmap has three parts:

• Keyboard model: you can specify your keyboard model with -model <YOUR-KEYBOARD-MODEL>.
• layout[s]: you can configure any number of available layouts for use with the keyboard. Note that you must use , to separate their names.
• Special options: you can specify special keyboard options with -options <YOUR-KEYBOARD-OPTIONS>.

In the command I use, I’ve set the keyboard model to thinkpad and configured both English and Persian.

In the special options section, using grp:shifts_toggle I’ve specified that switching between keyboard languages is done by pressing both Shift keys simultaneously.

It may seem unusual, but for me — since my hands are almost always on the keyboard — it’s more convenient than using Alt+Shift. I also have other keys for switching keyboard language so I can switch with one hand if needed.

With caps:escape_shifted_capslock I’ve specified that the CapsLock key acts as the Escape key.

I use the Vim editor for editing text files and frequently use the Escape key to exit various modes. The standard location of this key at the top-left corner of the keyboard is far from reach. Mapping CapsLock as Escape makes it much easier for me to use.

The next option is altwin:prtsc_rwin which specifies that the PrintScreen (PrtSc) key — located next to the right Alt key on my laptop keyboard — acts as a second Meta or Mod key (the Win logo key).

For years I have not used desktop environments on my computer, preferring window managers instead, which are lighter and more customizable. Most of my keyboard shortcuts are in the form Mod+<SOME-KEY>, and there are many of them. Having the Mod key on both sides of the keyboard makes using these shortcuts more comfortable.

Finally, lv3:ralt_switch specifies that layer 3 of the keyboard can be accessed using the right Alt key. As you know, each language on the keyboard has more than one layer of characters. In almost all cases, you can type layer 2 characters using the Shift key. For English these are uppercase letters and some other commonly used characters. For Persian on the standard keyboard, ژ is on the second layer of the ز key. Some other common characters are also on this layer, including ZWNJ (zero-width non-joiner). Interestingly, some less-used characters are on layer 3 and below. NBSP (non-breaking space) is one of them, located on layer 3 of the space key. I use the right Alt key to access this layer.

An important question is where to find a list of models, layouts, and special options. Simply open the documentation with man xkeyboard-config in your terminal and read it. You’ll find a comprehensive list of all options with clear explanations for each. Searching man xkeyboard-config <YOUR-DISTRIBUTION> in your browser will also find the documentation for your distribution. For example, for Arch Linux.

Of course, to find layouts you can also look in /usr/share/X11/xkb/symbols.

$ ls /usr/share/X11/xkb/symbols

The next question is how to make these settings permanent. I use the xorg-xinit package to start the X11 graphical environment and run the startx command directly from the console without using a Display Manager. When this command is run, it reads graphical environment settings from text files. One of these files is .xinitrc in the $HOME directory. I simply put any command I need to set up my graphical environment in a file and call it from inside .xinitrc. I’ve placed these commands in the .xprofile file in the $HOME directory.

$ cat ~/.xinitrc
#!/bin/sh

# xinitrc runs automatically when you run startx.

# There are some small but important commands that need to be run when we start
# the graphical environment. I keep those commands in ~/.xprofile because that
# file is run automatically if someone uses a display manager (login screen)
# and so they are needed there. To prevent doubling up commands, I source them
# here with the line below.

[ -f ~/.xprofile ] && . ~/.xprofile

# Fix Gnome Apps Slow  Start due to failing services
# Add this when you include flatpak in your system
dbus-update-activation-environment --systemd DBUS_SESSION_BUS_ADDRESS DISPLAY XAUTHORITY

# Here we start dwm.
# The loop is just to enable dwm's "restart" feature (mod+F2).
#exec dwm
ssh-agent dwm

$ cat ~/.xprofile
#!/usr/bin/sh

xset r rate 300 50 &    # Speed xrate up
unclutter &     # Remove mouse when idle
#xcompmgr &     # xcompmgr for transparency
picom --config ~/.config/picom/picom.conf &
dunst &                 # dunst for notifications
#keymap &       # some keyboard remap
mpd &
xrdb ~/.Xdefaults &
setbg &
setxkbmap -model thinkpad us,ir -option "grp:shifts_toggle,caps:escape_shifted_capslock,altwin:prtsc_rwin,lv3:ralt_switch" &

As you can see, the keyboard settings are specified in the last line of the file. Each Display Manager has configuration files that it calls before the graphical environment starts. You can find the path to these files by searching online and place the keyboard settings in them so they’re applied automatically.

xmodmap

xmodmap is a tool for editing key layouts on the keyboard in X11. You may need to swap two keys on the keyboard and that option may not be available among setxkbmap choices. For this, you can use xmodmap:

$ xmodmap -e 'keycode 135 = Super_R'

The above command converts the Menu key to the Super key, similar to what setxkbmap -option altwin:menu_win does. To find the code for each physical key on the keyboard, you can view the current keyboard layout list using xmodmap -pk.

$ xmodmap -pk
There are 10 KeySyms per KeyCode; KeyCodes range from 8 to 255.

    KeyCode Keysym (Keysym) ...
    Value   Value   (Name)  ...

      8
      9     0xff1b (Escape) ...
     10     0x0031 (1)  0x0021 (exclam) 0x10006f1 (Farsi_1) ...
...

The xmodmap -pke command gives you this list in a format usable by xmodmap.

A list of key names usable with xmodmap can be found on this page.

To find the code for a specific key, you can use xev. Simply run xev in the terminal and press the desired key to see its data.

$ xev | awk -F'[ )]+' '/^KeyPress/ { a[NR+2] } NR in a { printf "%-3s %s\n", $5, $8 }'
44  j
45  k
133 Super_L

There are many tools for working with the keyboard in X11 server. For example, xset -q gives you information about the current configuration state:

$ xset -q
Keyboard Control:
  auto repeat:  on    key click percent:  0    LED mask:  00000000
  XKB indicators:
    00: Caps Lock:   off    01: Num Lock:    off    02: Scroll Lock: off
  auto repeat delay:  300    repeat rate:  50
...

You can use it to check key states in scripts and configurations. You can also use xdotool to set the state of keyboard and mouse keys and the cursor position. xdotool is a tool for simulating input events in the X11 server environment.

$ xset -q | grep "Caps Lock:\s*on" && xdotool key Caps_Lock

The above command checks the state of CapsLock and disables it if it is active.

Swaywm on Wayland

Keyboard configuration in the Wayland environment is the responsibility of the compositor, and there is no single tool or uniform method for doing it. Compositors on Wayland receive inputs without an intermediary layer and get data at a lower level from the kernel’s APIs.

Keyboard configuration in Swaywm is done in this window manager’s configuration file. By default this file is at /etc/sway/config. If a configuration file exists at ~/.config/sway/config, that file takes higher priority.

My keyboard configuration for Swaywm in this file looks like this:

# Use `man xkeyboard-config` to view the options and `setxkbmap -query|print` to check the model
input {
    type:keyboard {
        xkb_layout "us,ir"
        xkb_model "thinkpad"
        xkb_options "caps:escape_shifted_capslock,grp:shifts_toggle,altwin:prtsc_rwin,lv3:ralt_switch"
    }

    type:touchpad {
        # no click needed
        tap enabled
        # disable touchpad while typing
        dwt enabled
        # natural_scroll enabled
        middle_emulation enabled
    }
}

Which are the same settings I had for the X11 server.

Tools like xmodmap, setxkbmap, xset, xev, and xdotool cannot be used on Wayland. The Swaywm wiki has a page about alternatives for common tools used in X11, which is a good guide. You can also visit this GitHub page to find more tools for Wayland.

Instead of xev you can use wev, and instead of xdotool you can use wtype, wlrctl, swaymsg seat <seat> cursor …, or ydotool. However, no replacement for xmodmap has been introduced — instead, a custom layout can be used.

Creating a Custom keymap

Even in the X11 environment, applying keyboard settings permanently using the tools we’ve discussed may not fully meet all your needs. Creating a keymap tailored to your personal needs gives you the ability to make any change to the character layout.

Before going further, let me describe the problem that prompted writing this post. A friend on Mastodon was looking for a way to replace / with the character ۀ on the Persian layout. This character is not on the standard Persian keyboard and is an Arabic character. Its correct form in Persian is written by combining the two characters ه and _ٔ. However, due to printing limitations, they were forced to use this character.

The best way to create a custom keymap is to use one of the existing files in /usr/share/X11/xkb/symbols and edit it.

For example, the file for the Persian keymap has this pattern:

$ cat /usr/share/X11/xkb/symbols/ir
// Iranian keyboard layout
...
default partial alphanumeric_keys
xkb_symbols "pes" {
    name[Group1]= "Persian";

    include "ir(pes_part_basic)"
    include "ir(pes_part_ext)"

    include "nbsp(zwnj2nb3nnb4)"
    include "level3(ralt_switch)"
};
...

In this file, the characters for each key are specified in an array corresponding to the different layers of that key. The row we’re interested in is key <AB10> { [ slash, Arabic_question_mark, question ] };. This means: pressing AB10 (the / key) types /; pressing it with Shift types ؟; and on layer 3 (with my configuration, using the right Alt key) it types ?.

To apply the change and create our custom keymap, we make a copy of this file and apply the changes to it. We can then use this custom layout.

The Unicode code for the character ۀ is U+06C0, which in this file should be written as 0x10006c0, following the same format as other Unicode characters:

$ sudo cp /usr/share/X11/xkb/symbols/{ir,ir.bak}

$ cat /usr/share/X11/xkb/symbols/ir | grep -n slash
89:    key <AB10> { [ slash,        Arabic_question_mark,   question    ] };
94:    key <BKSL> { [ backslash,        bar,            0x1002010   ] };

$ sudo sed -i 's/\[ slash,/\[ 0x10006c0,/' /usr/share/X11/xkb/symbols/my-ir

$ cat /usr/share/X11/xkb/symbols/ir | grep -n 0x10006c0
89:    key <AB10> { [ 0x10006c0,        Arabic_question_mark,   question    ] };

To find the name or code of Unicode characters, you can visit this website.

Although editing files in /usr/share/X11/xkb/symbols allows you to customize your keymap, since libxkbcommon 0.10.0 it has been possible to create personal configuration files in $XDG_CONFIG_HOME/xkb and in the folders $XDG_CONFIG_HOME/xkb/rules/ and $XDG_CONFIG_HOME/xkb/symbols. Files in this path are loaded before the default files.

For example, to add an option to replace the PrtSc key with the Menu key, we create and configure the files $XDG_CONFIG_HOME/xkb/rules/evdev and $XDG_CONFIG_HOME/xkb/symbols/custom as follows:

$ mkdir -p ~/.config/xkb/{symbols,rules}
$ cd ~/.config/xkb
$ cat > rules/evdev << EOF
! option = symbols
  custom:prtscmenu = +custom(prtscmenu)

! include %S/evdev
EOF
$ cat > symbols/custom << EOF
partial modifier_keys
xkb_symbols "prtscmenu" {
  key <PRSC> { [ Menu ] };
};
EOF

Then we need to enable/disable this option. For GNOME:

# apply custom layout:
$ gsettings set org.gnome.desktop.input-sources xkb-options "['custom:prtscmenu']"
# disable custom layout:
$ gsettings set org.gnome.desktop.input-sources xkb-options "[]"

Or to create the custom layout we made earlier:

$ cat > ~/.config/xkb/symbols/ir-mz << EOF
default partial alphanumeric_keys
xkb_symbols "basic" {
    include "ir(pes)"
    include "level3(caps_switch)"
    name[Group1] = "Persian (IR, my custom)";
    key <AB10> { [ 0x10006c0, Arabic_question_mark, question ] };
};
EOF

I use the following file to swap the چ character (which is less commonly needed in certain workflows) with / in the standard Persian keyboard:

$ cat ~/.config/xkb/symbols/ir-mz
default partial alphanumeric_keys
xkb_symbols "basic" {
    include "ir(pes)"
    name[Group1]= "Persian (IR, my custom layout)";
    key <AB10> { [ Arabic_tcheh,		Arabic_question_mark,	question	] };
    key <AD12> { [ slash,	braceleft,		0x100202b	] };
};

To use this layout in Swaywm, add input type:keyboard xkb_layout ir-mz to the configuration file, or use this command:

swaymsg input "$(swaymsg -t get_inputs | grep 'identifier.*keyboard' | cut -d'"' -f4)" xkb_layout ir-mz

As mentioned, these settings are loaded and applied before the default paths. For more information, you can read this page.

I use the foot terminal emulator on my system, which runs correctly on wayland and is very simple, fast, and minimal. For transferring files between my Android phone and my GNU/Linux system, I use scp or rsync over ssh. If you’re interested, you can watch the tutorial video I made about this.

When I first SSH’d from the foot terminal into my Android phone, the terminal colors were not displayed correctly and every character I typed appeared twice. After a bit of searching, I found that the problem was related to missing terminfo data on termux. Here you can see the missing colors and the repeated zsh prompt data on Termux after SSH, and compare it with the phone:

Missing terminfo Data on the Server

You may have encountered these errors many times when running a terminal-based program on a server you’ve SSH’d into:

Error opening terminal: xterm-kitty.
# OR
WARNING: terminal is not fully functional
/etc/hosts  (press RETURN)

These problems occur for the same reason I mentioned.

What to Do?

• On the server, change the $TERM environment variable with export TERM=xterm and test whether the problem is resolved.
• Install the terminal emulator you’re using on the server as well. (For me this was not possible since I wanted to SSH from Termux.)
• Add the terminfo data to the server.

Adding terminfo Data to the Server

• If you have the terminfo data in a file, you can transfer it to the server and install it using tic.

rsync <PATH-TO-YOUR.TERMINFO> <REMOTE-MACHINE>:/tmp
ssh <REMOTE-MACHINE>
sudo tic /tmp/emu.terminfo

If you don’t have the data, you can view it and save it to a file using infocmp > foot.terminfo. For foot, the data looks like this:

infocmp
#	Reconstructed via infocmp from file: /usr/share/terminfo/f/foot
foot|foot terminal emulator,
	am, bce, bw, ccc, hs, mir, msgr, npc, xenl,
	colors#0x100, cols#80, it#8, lines#24, pairs#0x10000,
	acsc=``aaffggiijjkkllmmnnooppqqrrssttuuvvwwxxyyzz{{||}}~~,
	bel=^G, blink=\E[5m, bold=\E[1m, cbt=\E[Z, civis=\E[?25l,
	clear=\E[H\E[2J, cnorm=\E[?12l\E[?25h, cr=\r,
	csr=\E[%i%p1%d;%p2%dr, cub=\E[%p1%dD, cub1=^H,
	cud=\E[%p1%dB, cud1=\n, cuf=\E[%p1%dC, cuf1=\E[C,
	cup=\E[%i%p1%d;%p2%dH, cuu=\E[%p1%dA, cuu1=\E[A,
	cvvis=\E[?12;25h, dch=\E[%p1%dP, dch1=\E[P, dim=\E[2m,
	dl=\E[%p1%dM, dl1=\E[M, ech=\E[%p1%dX, ed=\E[J, el=\E[K,
	el1=\E[1K, flash=\E]555\E\\, home=\E[H, hpa=\E[%i%p1%dG,
	ht=^I, hts=\EH, ich=\E[%p1%d@, ich1=\E[@, il=\E[%p1%dL,
	il1=\E[L, ind=\n, indn=\E[%p1%dS,
	initc=\E]4;%p1%d;rgb:%p2%{255}%*%{1000}%/%2.2X/%p3%{255}%*%{1000}%/%2.2X/%p4%{255}%*%{1000}%/%2.2X\E\\,
	invis=\E[8m, is2=\E[!p\E[?3;4l\E[4l\E>, kDC=\E[3;2~,
	kEND=\E[1;2F, kHOM=\E[1;2H, kIC=\E[2;2~, kLFT=\E[1;2D,
	kNXT=\E[6;2~, kPRV=\E[5;2~, kRIT=\E[1;2C, kbs=^?,
	kcbt=\E[Z, kcub1=\EOD, kcud1=\EOB, kcuf1=\EOC, kcuu1=\EOA,
	kdch1=\E[3~, kend=\EOF, kf1=\EOP, kf10=\E[21~, kf11=\E[23~,
	kf12=\E[24~, kf13=\E[1;2P, kf14=\E[1;2Q, kf15=\E[1;2R,
	kf16=\E[1;2S, kf17=\E[15;2~, kf18=\E[17;2~,
	kf19=\E[18;2~, kf2=\EOQ, kf20=\E[19;2~, kf21=\E[20;2~,
	kf22=\E[21;2~, kf23=\E[23;2~, kf24=\E[24;2~,
	kf25=\E[1;5P, kf26=\E[1;5Q, kf27=\E[1;5R, kf28=\E[1;5S,
	kf29=\E[15;5~, kf3=\EOR, kf30=\E[17;5~, kf31=\E[18;5~,
	kf32=\E[19;5~, kf33=\E[20;5~, kf34=\E[21;5~,
	kf35=\E[23;5~, kf36=\E[24;5~, kf37=\E[1;6P, kf38=\E[1;6Q,
	kf39=\E[1;6R, kf4=\EOS, kf40=\E[1;6S, kf41=\E[15;6~,
	kf42=\E[17;6~, kf43=\E[18;6~, kf44=\E[19;6~,
	kf45=\E[20;6~, kf46=\E[21;6~, kf47=\E[23;6~,
	kf48=\E[24;6~, kf49=\E[1;3P, kf5=\E[15~, kf50=\E[1;3Q,
	kf51=\E[1;3R, kf52=\E[1;3S, kf53=\E[15;3~, kf54=\E[17;3~,
	kf55=\E[18;3~, kf56=\E[19;3~, kf57=\E[20;3~,
	kf58=\E[21;3~, kf59=\E[23;3~, kf6=\E[17~, kf60=\E[24;3~,
	kf61=\E[1;4P, kf62=\E[1;4Q, kf63=\E[1;4R, kf7=\E[18~,
	kf8=\E[19~, kf9=\E[20~, khome=\EOH, kich1=\E[2~,
	kind=\E[1;2B, kmous=\E[<, knp=\E[6~, kpp=\E[5~,
	kri=\E[1;2A, oc=\E]104\E\\, op=\E[39;49m, rc=\E8,
	rep=%p1%c\E[%p2%{1}%-%db, rev=\E[7m, ri=\EM,
	rin=\E[%p1%dT, ritm=\E[23m, rmacs=\E(B, rmam=\E[?7l,
	rmcup=\E[?1049l\E[23;0;0t, rmir=\E[4l, rmkx=\E[?1l\E>,
	rmso=\E[27m, rmul=\E[24m, rs1=\Ec,
	rs2=\E[!p\E[?3;4l\E[4l\E>, sc=\E7,
	setab=\E[%?%p1%{8}%<%t4%p1%d%e%p1%{16}%<%t10%p1%{8}%-%d%e48:5:%p1%d%;m,
	setaf=\E[%?%p1%{8}%<%t3%p1%d%e%p1%{16}%<%t9%p1%{8}%-%d%e38:5:%p1%d%;m,
	sgr=%?%p9%t\E(0%e\E(B%;\E[0%?%p6%t;1%;%?%p5%t;2%;%?%p2%t;4%;%?%p1%p3%|%t;7%;%?%p4%t;5%;%?%p7%t;8%;m,
	sgr0=\E(B\E[m, sitm=\E[3m, smacs=\E(0, smam=\E[?7h,
	smcup=\E[?1049h\E[22;0;0t, smir=\E[4h, smkx=\E[?1h\E=,
	smso=\E[7m, smul=\E[4m, tbc=\E[3g, u6=\E[%i%d;%dR,
	u7=\E[6n, u8=\E[?%[;0123456789]c, u9=\E[c,
	vpa=\E[%i%p1%dd,

• The simpler way is to generate the data and pipe it directly to the server for installation:

infocmp | ssh <REMOTE-MACHINE> "tic -x /dev/stdin"

• But if your server is old or for some reason tic cannot access /dev/stdin, you can write the data directly to a file on the server and then install it:

infocmp | ssh <REMOTE-MACHINE> "cat > /tmp/terminfo && tic -x /tmp/terminfo; rm /tmp/terminfo"

After adding foot data to termux, my problem was resolved.

Resources:

• kitty FAQ
• foot repository

A friend of mine has a Lenovo ThinkPad 10 tablet and, wanting to install GNU/Linux on it, asked for my help. I was also curious to see how well different desktop environments work with touchscreens.

To avoid having to reformat a USB drive every time you want to boot a live ISO, you can use ventoy.

After testing several options, we decided to install Ubuntu. Since it was my friend’s first time using GNU/Linux, we decided to go step by step and I would explain the process at each stage. But right at the start, things got stuck. I tried several times using Rufus to create a bootable USB from the ISO file. Even though I had set USB as the first boot priority in BIOS settings (accessible by pressing the power button and volume-up simultaneously, where you can also change boot settings), disabled Secure Boot and Fast Boot — the system kept booting from the hard drive and wouldn’t recognize the USB drive as bootable.

I also tried the familiar method of writing the ISO to USB using dd, but while it booted fine on a laptop, it had problems on the tablet.

dd if=<PATH-TO-A-LIVE-ISO> of=/dev/sdx status="progress"

Here sdx is the block device corresponding to your USB drive, which you can find using lsblk.

After a bit of searching and checking the tablet’s specs, I realized this system uses a 32-bit UEFI, which is exactly why the USB was not booting.

What to Do?

The answer is simple:

• Format the USB drive as fat32.
• Manually copy the contents of the ISO file onto it.

To mount an ISO file on GNU/Linux, use the following command:

sudo mount -o loop <PATH-TO-A-LIVE-ISO> <PATH-TO-THE-FOLDER-YOU-WANT-TO-MOUNT-THE-ISO>

• Download the bootia32.efi file from the GitHub repository or from here and copy it to the /EFI/BOOT folder.
• Copy the grub.cfg file from the /boot/grub folder and place it in the root of the USB (/).
• Disable Fast Boot and Secure Boot, then boot the system from the USB.

انگار می توانست همه آنچه را که ته باغ بالای چاه کنار اتاقک در کله پیرمرد گذشته بود بخواند. به خاطرش آمد نوه‌اش هم همانطور خیره نگاهش کرده بود. معلوم بود پسرش همان وقت تصمیم نگرفته بود. خوب که نگاه کرد مطمئن شد پسرش همان چشم‌ها را داشت. همان‌ها که پرهیز می‌کرد از دیدنشان در آینه. همان چشم‌های مریم که هر بار به آینه نگاه می‌کرد، با غضب نگاهش می‌کردند. مریم مادرش بود انگار همه این اتفاق‌ها را او سال‌ها پیش رقم زده بود با یک نفرین. همان موقع که روی تخت افتاده بود و بعد از بیست سال به هوای اینکه می‌خواست توی ده بمیرد پسرش را کشانده بود آنجا. زن سوم علیخان بود بعد از دو تا زن که هیچ کدام بچه نیاورده بودند. همه می‌گفتند اجاق علیخان کور است و بی وارث می‌ماند. علیخان چهل و پنج سالی داشت که مریم را گرفته بود. همان‌جا توی ده آبستن شده بود و تنها پسر علیخان را آورده بود.

مریم

آسمان ابر گرفته و هوا نمدار بود. مردها همگی رفته بودند دنبال گرگ‌ها، که آن سال بد جوری جری شده بودند و تا پای آغل‌ها هم آمده بودند. نه پیش از آن سال دیده بودش و نه بعد از آن. حالا بعد از بیست سال روی تخت افتاده بود و می‌دانست که امشب را به صبح نمی‌رساند. صدای جیرجیرک‌ها را می‌شنید و بوی نم باغچه را، که باد همراه عطرچوب‌های تازه و گل‌هایی که نمی‌شناخت می‌آورد. تنها چیزی که می‌خواست این بود که یک بار دیگر قیافه سلیمان را ببیند و بی هیچ بهانه آوردنی همه چیز را برای خداداد گفته بود.

خداداد

خداداد آشفته پالتویش را برداشت و از در جلوی باغچه زد بیرون. تا میدانگاهی هیچ کس نبود. بی معطلی راه افتاد. می‌خواست تا از خانه اربابی و همه آنچه شنیده بود بگریزد. باد بوته چرخه ای را توی کوچه خالی می‌غلتاند. پایش توی گل کنار جو رفت و سر خورد. بلند شد و لباسش را نگاه کرد، تا زیر سینه‌اش گل شده بود. دور میدانگاهی یک قهوه خانه بود و خوار و بار فروشی که همه احتیاحات ابتدایی اهالی را رفع می‌کرد، نانوایی و عطاری که قیافه‌اش را توی شیشه آن نگاه کرد. لب‌های باریک و کشیده، چانه تیز، موهای بورش و با خودش فکر کرد: «غیر از چشم‌هام». یاد سلیمان افتاد و لرزش گرفت. راهش را ادامه داد. جلوی قهوه خانه داشتند خون‌های گوسفند را می‌شستند. سگ سیاهی با گوش‌های بریده شده آشغال‌های لاشه را به دندان می‌کشید. پسر بچه‌ای هم با لباس قرمز آستین پاره و کله از ته تراشیده چمباتمه زده بود. صورتی کثیف و آفتاب سوخته داشت. چشم‌هایش با نگاهی حریص سگ را می پایید. همه اینها را خوب به یاد داشت. و اینکه بعد از آن پیچید توی یک کوچه خلوت و کنار چنار خشکی به دیوار تکیه داد. پوستش رفته بود و تنه‌اش پوک شده بود. با این حال با سماجت سر پا مانده بود. آتش سیگار که به انگشت‌هاش رسید و سوزاندش، باعث شد به خودش بیاید. هیچ جوری توی کتش نمی‌رفت. او پسر علیخان بود و البته وارثش و حالا همه کاره ملک اربابی. با قدم‌های تند راه برگشت را پیش گرفت. به خانه که رسید دسته بیل بلندی را برداشت و راهش را به طرف اتاقک ته باغ کج کرد.

سلیمان

آتش ذغال‌هایش توی ایوان از بالای کاهدان انبار معلوم بود. علیخان نشسته بود و مثل هر شب وافور را گذاشته بود گوشه لبش. گویی عزمش را جزم کرده بود قبل از مرگ هر چه را که داشت دود کند. بغل ملا نشسته بود و آتش ذغال را فوت می‌کرد. گر می‌گرفت و تا جرقه میزد، شروع می‌کرد به فحش دادن که: «قرمساق‌ها مگر نگفتم خوب بگردانیدش.» اهل نماز و روزه بود، ولی حتی ماه رمضان هم نوبت‌های تریاکش ترک نمی‌شد. مجوزش را هم از ملا گرفته بود. البته سهم ملا هم از همنشینی پای بساط می‌رسید، که هیچ وقت از تو نمی‌انداخت. حتی آن شب هم که همه رفته بودند دنبال گرگ‌ها. اول از همه او دیده بودشان، چشم‌هایشان را که برق می‌زد. تیر انداخته بود از همان بالای آغل که نشسته بود. خیالش آمده بود سگند، قبل از اینکه خوب نزدیک شوند. به همین خاطر تیرش نشسته بود روی پشت یکیشان. سگ‌ها هم معلوم نبود صدایشان چرا در نیامده بود. حالا هم که رفته بودند رد بزنند، سلیمان مانده بود بالای کاهدان بپا. صدا که از پایین آمد خشکش زد. فقط او می‌دانست چون خودش آنجا بود. اما بقیه می‌گفتند بعد از اینکه پدرش داده بودش به علیخان نفرین کرده بود و پدرش از سر همان نفرین مرده بود. سلیمان اما می‌دانست مریم کار را فقط به نفرین نسپرده بود. هر شب می‌دیدش که تا کنار جو می‌رفت و پاهایش را توی آب می‌گذاشت. اما امشب فرق می‌کرد و تا در انبار را باز نکرد، متوجه‌اش نشد. موهای بلندش از پشت روسری گلدار قرمزی که مثل دستمال بسته بود بالای سرش بیرون ریخته بود. یک طره مو بر پیشانی خیسش افتاده بود و روی شقیقه از کنار چشم راستش تا روی گوشش کشیده شده بود. چند تار مو هم بر گونه‌های تبدارش خزیده بود و افشان شده بود. چشم و ابروی مشکی داشت. در چشم‌های درشتش رگ‌های سرخ دویده بود، لبریز از شرری مفتون کننده. لب‌هایش گلگون نبود ولی او دوست داشت. صورتی رنگ پریده بود و شیارهایشان از خشکی نبود. یک جوری انگار انار رسیده که ترک می‌خورد. قطره‌ای عرق از زیر چانه بر گردن برهنه کنار رگ نزدیک گلویش لغزید. خرده‌های ریز کاه بر گودی استخوان‌های ترقوه‌اش برق می‌زد، زیر نور مرده مهتاب که به زحمت از شکاف‌های سقف بر بدنش می‌نشست. شال سفیدی روی دوشش انداخته بود و زیر پستان‌هایش گره زده بود. لباسش تر شده بود و نفس نفس که می‌زد پستان‌هایش بالا و پایین می‌شد، بی‌فاصله چسبیده به پیرهنی که دور کمر باریکش چین خورده بود. بر ساق و مچ سفیدش هیچ النگویی نبود. لبه پایین دامنش را بالا گرفته بود و پاهایش تا زانو پیدا بود. پاشنه پایش را بالا گرفته بود. ماهیچه‌های پشت ساق‌های باریک و ورزیده‌اش منقبض و منبسط می‌شد. بر کف خاکی انبار که قدم میزد پاشنه و گودی پشت زانویش پر و خالی می‌شد و انحنای روی رانش از زیر دامن بالا می‌آمد و برجسته می‌شد. نگاهی به دور و برش انداخت و آرام سلیمان را صدا زد. سلیمان نفسش را که حبس شده بود بیرون داد و از روی کاهدان پایین آمد. خیره به ذرات غبار که در ستون‌های نور ملایم اطراف او می‌چرخید نگاه کرد. دستهایش را گرفت گونه‌اش را بوسید و او را به بغل گرفت. کنار هم روی علف‌های تازه دراز کشیدند. نمی‌دانست چقدر از این‌ها را مریم برای خداداد گفته بود. توی بغل سلیمان که دراز کشیده بود، لختی اندام‌هاش رنگ پریده می‌نمود. به جز گونه‌ها و خیرگی چشم‌هاش که هراسی بی حیا ته آنها موج می‌زد. اصلا مگر هیچ کس غیر از سلیمان که این تصویر را تا آن لحظه مثل یک راز با خودش حمل کرده بود، می‌توانست این معجزه را تعریف کند. چشم‌های خداداد چقدر شبیه چشم‌های مریم بود، وقتی که پدرش را برای آخرین بار نگاه می‌کرد.

خداداد

مریم برایشان فقط چشم‌هایش را به ارث نگذاشته بود. اینها را حالا خداداد می‌دید که با سری شکافته به چشم‌های پسرش خیره شده یود. به خاطرش آمد نوه‌اش هم با آن لباس قرمز آستین پاره و کله کچلش همان‌طور خیره نگاهش کرده بود.

From Wikipedia, the free encyclopedia

Mastodon is a free and open-source self-hosted social networking service that allows anyone to host their own server node in the network, with user bases distributed across different servers.

These servers are interconnected as a federated social network, allowing their users to interact seamlessly with one another. Mastodon is part of a larger federation that enables users to interact with other free platforms such as PeerTube and Friendica that support the same protocols.

Adding a Comments Section with Mastodon

One of the challenges with static websites is adding a comments section. There are various options for doing this. One simple solution is Disqus. However, Disqus was never an appealing option for me because it is heavy, loads a lot of unnecessary scripts on pages, and most importantly is not free software. For more alternatives, you can see this post by Mehdi Sadeghi.

Thanks to ‪@carl‬ for writing the code and the report in this post, you can now send me your comments using a Mastodon account, and view what others have written.

JavaScript Code

Since the original code was written for #HUGO, I made minor modifications to make it compatible with #JEKYLL and placed it in the mastodon.html file inside the _includes folder:

<div class="page-content">
  <h2>Comments</h2>
  <p>You can view the comments for this post on Mastodon and <a class="link" href="https://{{ page.mastodon.host }}/@{{ page.mastodon.username }}/{{ page.mastodon.id }}" target="_blank" rel="noopener noreferrer"> here</a>. To write your own comment, click the link below.</p>
  <p><a class="button" href="https://{{ page.mastodon.host }}/@{{ page.mastodon.username }}/{{ page.mastodon.id }}" target="_blank" rel="noopener noreferrer">Write a comment</a></p>
  <p></p>
  <p id="mastodon-comments-list"><a id="load-comment" class="button">View comments</a></p>
  <div id="comments-wrapper">
    <noscript><p>JavaScript must be enabled for this section, or view the <a href="https://{{ page.mastodon.host }}/@{{ page.mastodon.username }}/{{ page.mastodon.id }}" target="_blank" rel="noopener noreferrer">original post</a> on Mastodon</p></noscript>
  </div>
  <script src="/assets/js/purify.min.js"></script>
  <script type="text/javascript">
    function escapeHtml(unsafe) {
      return unsafe
        .replace(/&/g, "&amp;")
        .replace(/</g, "&lt;")
        .replace(/>/g, "&gt;")
        .replace(/"/g, "&quot;")
        .replace(/'/g, "&#039;");
    }

    function emojify(input, emojis) {
      let output = input;

      emojis.forEach(emoji => {
        let picture = document.createElement("picture");

        let source = document.createElement("source");
        source.setAttribute("srcset", escapeHtml(emoji.url));
        source.setAttribute("media", "(prefers-reduced-motion: no-preference)");

        let img = document.createElement("img");
        img.className = "emoji";
        img.setAttribute("src", escapeHtml(emoji.static_url));
        img.setAttribute("alt", `:${ emoji.shortcode }:`);
        img.setAttribute("title", `:${ emoji.shortcode }:`);
        img.setAttribute("width", "20");
        img.setAttribute("height", "20");

        picture.appendChild(source);
        picture.appendChild(img);

        output = output.replace(`:${ emoji.shortcode }:`, picture.outerHTML);
      });

      return output;
    }

    function loadcomments() {
      document.getElementById("load-comment").innerHTML = "Loading comments";
      fetch('https://{{ page.mastodon.host }}/api/v1/statuses/{{ page.mastodon.id }}/context')
        .then(function(response) {
          return response.json();
        })
        .then(function(data) {
          let descendants = data['descendants'];
          if(
            descendants &&
            Array.isArray(descendants) &&
            descendants.length > 0
          ) {
            document.getElementById('mastodon-comments-list').innerHTML = "";
            descendants.forEach(function(reply) {
              if( reply.account.display_name.length > 0 ) {
                reply.account.display_name = escapeHtml(reply.account.display_name);
                reply.account.display_name = emojify(reply.account.display_name, reply.account.emojis);
              } else {
                reply.account.display_name = reply.account.username;
              };

              reply.content = emojify(reply.content, reply.emojis);

              reply.created_at = new Date( reply.created_at ).toLocaleString('en-US', {
                dateStyle: "long",
                timeStyle: "short",
              });

              let instance = "";
              if( reply.account.acct.includes("@") ) {
                instance = reply.account.acct.split("@")[1];
              } else {
                instance = "{{ page.mastodon.host }}";
              }

              mastodonComment =
                `<div class="mastodon-comment">
                <div class="avatar">
                <img src="${escapeHtml(reply.account.avatar_static)}" height=60 width=60 alt="">
                </div>
                <div class="content">
                <div class="author" title="profile: @${ reply.account.username }@${ instance }">
                <a href="${reply.account.url}" target="_blank" rel="external nofollow">
                <span>${reply.account.display_name}</span>
                <span class="instance">${instance}</span>
                </a>
                </div>
                <div title="View comment on ${ instance }">
                <a class="date" href="${reply.uri}" target="_blank" rel="external nofollow">
                ${reply.created_at}
                </a>
                </div>
                <div class="mastodon-comment-content">${reply.content}</div>
                </div>
                </div>`;
              document.getElementById('mastodon-comments-list').appendChild(DOMPurify.sanitize(mastodonComment, {'RETURN_DOM_FRAGMENT': true}));
            });
          } else {
            document.getElementById('mastodon-comments-list').innerHTML = "<p>No comments yet</p>";
          }
        });
    }
    document.getElementById("load-comment").addEventListener("click", loadcomments);
  </script>
</div>

Then in the post layout, I added a condition so that whenever mastodon is present in the post’s frontmatter, the comments section is included:

{% if page.mastodon %}
  {% include mastodon.html %}
{% endif %}

And in the frontmatter, mastodon should be added with this structure:

mastodon:
  host: mas.to
  username: mz
  id: 105582586560918183

Where the following must be specified:

• host: the Mastodon server address
• username: the Mastodon username
• id: the unique ID of the toot related to the post

Finally, I used these few lines of CSS for better comment display.

You can try the final result here:

SoX is a tool for editing audio files, available on GNU/Linux, macOS, and Windows. You can also use it on Android inside the termux environment.

In this post, we will first use SoX to reduce noise in an audio file, and then examine a script for noise reduction in audio and video files using SoX and FFmpeg.

Noise Reduction with SoX

If we have an audio file called audio.wav that has a small section of silence — or more accurately, ambient noise — recorded at the beginning, we can extract that section using SoX:

# sox in.ext out.ext trim {start: s.ms} {duration: s.ms}
sox audio.wav noise.wav trim 0 0.900

Here, SoX takes audio.wav as input, extracts the first 0.9 seconds of it, and saves it as noise.wav.

Now we use this section to create a noise profile file called noise.prof:

sox noise.wav -n noiseprof noise.prof

Finally, we perform noise reduction based on the generated profile. A sensitivity factor is used to control the amount of noise reduction; according to this source, a value between 0.2 and 0.3 produces the best results:

sox audio.wav audio-clean.wav noisered noise.prof 0.21

A Script for Noise Reduction in Audio and Video Files with FFmpeg and SoX

This script takes two paths — first the input file, then the output file ($1 and $2). If the input file is a video, it uses FFmpeg to extract the audio. It then reduces noise in the audio file as described above, and finally, if the input was a video, uses FFmpeg again to merge the audio and video back together.

#!/usr/bin/sh

usage ()
{
    printf "Usage : noisereduce <input video file> <output video file>\n"
    exit
}

# Tests for requirements
ffmpeg -version >/dev/null || { echo >&2 "We require 'ffmpeg' but it's not installed. Install it by 'sudo apt-get install ffmpeg' Aborting."; exit 1; }
sox --version >/dev/null || { echo >&2 "We require 'sox' but it's not installed. Install it by 'sudo apt-get install sox' Aborting."; exit 1; }

if [ "$#" -ne 2 ]
then
  usage
fi

if [ ! -e "$1" ]
then
    printf "File not found: %s\n" "$1"
    exit
fi

if [ -e "$2" ]
then
    printf "File %s already exists, overwrite? [y/N]\n: " "$2"
    read -r yn
    case $yn in
        [Yy]* ) ;;
        * ) exit;;
    esac
fi

inBasename=$(basename "$1")
inExt="${inBasename##*.}"

isVideoStr=$(ffprobe -v warning -show_streams "$1" | grep codec_type=video)
if [ -n "$isVideoStr" ]
then
    isVideo=1
    printf "Detected %s as a video file\n" "$inBasename"
else
    isVideo=0
    printf "Detected %s as an audio file\n" "$inBasename"
fi

printf "Sample noise start time [00:00:00]: "
read -r sampleStart
if [ -z "$sampleStart" ] ; then sampleStart="00:00:00"; fi
printf "Sample noise end time [00:00:00.900]: "
read -r sampleEnd
if [ -z "$sampleEnd" ] ; then sampleEnd="00:00:00.900"; fi
printf "Noise reduction amount [0.21]: "
read -r sensitivity
if [ -z "$sensitivity" ] ; then sensitivity="0.21"; fi


tmpVidFile="/tmp/noiseclean_tmpvid.$inExt"
tmpAudFile="/tmp/noiseclean_tmpaud.wav"
noiseAudFile="/tmp/noiseclean_noiseaud.wav"
noiseProfFile="/tmp/noiseclean_noise.prof"
tmpAudCleanFile="/tmp/noiseclean_tmpaud-clean.wav"

printf "Cleaning noise on %s...\n" "$1"

if [ $isVideo -eq "1" ]; then
    ffmpeg -v warning -y -i "$1" -qscale:v 0 -vcodec copy -an "$tmpVidFile"
    ffmpeg -v warning -y -i "$1" -qscale:a 0 "$tmpAudFile"
else
    cp "$1" "$tmpAudFile"
fi
ffmpeg -v warning -y -i "$1" -vn -ss "$sampleStart" -t "$sampleEnd" "$noiseAudFile"
sox "$noiseAudFile" -n noiseprof "$noiseProfFile"
sox "$tmpAudFile" "$tmpAudCleanFile" noisered "$noiseProfFile" "$sensitivity"
if [ $isVideo -eq "1" ]; then
    ffmpeg -v warning -y -i "$tmpAudCleanFile" -i "$tmpVidFile" -vcodec copy -qscale:v 0 -qscale:a 0 "$2"
else
    cp "$tmpAudCleanFile" "$2"
fi

printf "Done"

More Links

• A few examples in the main repository
• SoX tricks
• A fork on GitHub
• Supported features and formats
• A C# example

If you prefer watching a tutorial video or want to hear the difference between the input and output:

Firefox Extensions

Before November 2017, various tools existed for building a Firefox extension, but in newer versions some of them — such as overlay add-ons, bootstrapped add-ons, Add-on SDK, and others — are no longer supported. New extensions must be built using WebExtensions APIs. If you have a Legacy extension and want to make it compatible with newer versions of Firefox, you can use this Mozilla guide.

Structure of WebExtensions

Every WebExtension has a manifest.json file that defines its structure, resources, and properties. The image below shows the general structure of the manifest.json file in a WebExtension.

Sample manifest.json file for the Mahour extension:

{
  "manifest_version": 2,
  "name": "Mahour",
  "description": "Adds a new Iranian(Persian/Jalali/Khorshidi) date column to ThunderBird.",
  "version": "1.1.2",
  "homepage_url": "https://github.com/mhdzli/mahour",
  "author": "M.Zeinali",
  "applications": {
    "gecko": {
      "id": "mahour@zmim.ir",
      "strict_min_version": "68.0a1"
    }
  },
  "[experiment_apis](experiment_apis)": {
    "MahourDate": {
      "schema": "api/schema.json",
      "parent": {
        "scopes": ["addon_parent"],
        "paths": [["MahourDate"]],
        "script": "api/experiments.js"
      }
    }
  },
  "background": {
    "scripts": ["background/background.js"]
  },
  "browser_action": {
    "default_title": "ماهور",
    "default_popup": "popup/popup.html",
    "default_icon": {
      "48": "assets/icons/icon48.png"
    }
  },
  "options_ui": {
    "page": "options/options.html",
    "open_in_tab": false
  },
  "icons": {
    "32": "assets/icons/icon32.png",
    "48": "assets/icons/icon48.png",
    "64": "assets/icons/icon64.png",
    "128": "assets/icons/icon128.png"
  },
  "permissions": ["storage"]
}

This file contains the following keys:

• Essential keys that must be defined:
  • Manifest version (manifest_version)
  • Extension name (name)
  • Extension version (version)
• Optional keys:
  • Application description (description)
  • Icons (icons)
  • Extension website or repository (homepage_url)
  • Author (author)
• Browser-specific settings (browser_specific_settings or applications here). Browser properties required to run the extension when needed are specified in this key:
  • Minimum version of the browser’s content rendering engine (gecko: content rendering engine)
  • Extension ID for the content engine (gecko.id)
• Background scripts (Background scripts): these scripts run in the background in a special page called the background page.
• Content scripts (Content scripts) that do the main work and generate content for different parts of the extension. Mahour uses experiment_apis instead of this key. The content script is loaded inside the API.
• Experimental API (experiment_apis): with this key, a new API can be created for use in the extension.
• Default pages that can use css and js files like a regular webpage:
  • Sidebars (Sidebars)
  • Popups — their properties are defined in the browser_action || page_action key.
  • Options: their properties are defined in the options_ui key. This key makes the Preferences option visible for the extension in the Add-ons Manager page. Clicking on it loads the options page in that section and allows changing extension properties. (Setting the open_in_tab property will open a new page for loading.)
• Extension pages (Extension pages): each extension can also contain its own custom pages in addition to the default pages, which are defined here.
• Web accessible resources (Web accessible resources): if we want to use HTML, CSS, JavaScript, etc. files in the extension’s content generation, we specify them here. (Example: if the extension needs to display images on web pages, we path them here so we can access them.)
• Permissions (permissions): the permissions required by the application are specified with this key. For optional permissions that won’t block execution if absent, the optional_permissions key is used.

Extension APIs

To better understand Mozilla’s APIs and how to use them, read this guide.

Below is an example of an API defined in manifest.json:

{
  "manifest_version": 2,
  "name": "Extension containing an experimental API",
  "experiment_apis": {
    "apiname": {
      "schema": "schema.json",
      "parent": {
        "scopes": ["addon_parent"],
        "paths": [["myapi"]],
        "script": "implementation.js"
      },

      "child": {
        "scopes": ["addon_child"],
        "paths": [["myapi"]],
        "script": "child-implementation.js"
      }
    }

}
}

In each API, one or more namespaces are defined, whose objects can be called in extension scripts. A schema must be defined for each API that specifies the API’s properties. The schema key in manifest.json specifies a relative path within the extension’s root for the schema file.

The main properties of the parent process and child processes are specified with the parent and child keys and their script paths.

Currently, the only allowed options for the scopes key for specifying the access scope of each namespace are addon-child and addon-parent.

The schema file for the Mahour extension:

[
  {
    "namespace": "MahourDate",
    "functions": [
      {
        "name": "addWindowListener",
        "type": "function",
        "description": "Adds a listener 3pane windows",
        "async": false,
        "parameters": [
          {
            "name": "hich",
            "type": "string",
            "description": "hich"
          }
        ]
      },
      {
        "name": "changeSettings",
        "type": "function",
        "description": "Handle Prefrences",
        "async": true,
        "parameters": [
          {
            "name": "newSettings",
            "type": "object",
            "properties": {
              "longMonth": { "type": "boolean" },
              "showTime": { "type": "boolean" },
              "weekDay": { "type": "boolean" },
              "englishNumbers": { "type": "boolean" }
            }
          }
        ]
      }
    ]
  }
]

In this file, two functions addWindowListener and changeSettings are defined for the MahourDate namespace. The first function adds a windowListener to control the display of the Date column at program startup. The second function rebuilds and displays the Date column when the settings are changed.

Note: Calling these functions in the background script must be accompanied by a variable. Therefore, even though addWindowListener doesn’t need any input, a variable named hich (meaning “nothing” 😂) is defined for it.

These functions are defined inside the api/experiments.js script, which is referenced in the parent property of the MahourDate API in the extension’s manifest.json file.

The experiment.js file:

// This Source Code Form is subject to the terms of the
// GNU General Public License, version 3.0.

"use strict";

var { Services } = ChromeUtils.import("resource://gre/modules/Services.jsm");
var { ExtensionSupport } = ChromeUtils.import(
"resource:///modules/ExtensionSupport.jsm"
);
var { ExtensionParent } = ChromeUtils.import(
"resource://gre/modules/ExtensionParent.jsm"
);

const EXTENSION_NAME = "mahour@zmim.ir";
var extension = ExtensionParent.GlobalManager.getExtension(EXTENSION_NAME);

//customizeable options
var monthStyle = "long";
var timeStyle = "2-digit";
var weekDayStyle = "hidden";
var numbersStyle = "arabext";

// Implements the functions defined in the experiments section of schema.json.
var MahourDate = class extends ExtensionCommon.ExtensionAPI {
onStartup() {}

onShutdown(isAppShutdown) {
if (isAppShutdown) return;
Services.obs.notifyObservers(null, "startupcache-invalidate");
}

getAPI(context) {
context.callOnClose(this);
return {
MahourDate: {
addWindowListener(hich) {
ExtensionSupport.registerWindowListener(EXTENSION_NAME, {
chromeURLs: [
"chrome://messenger/content/messenger.xul",
"chrome://messenger/content/messenger.xhtml",
],
onLoadWindow: paint,
onUnloadWindow: unpaint,
});
},
changeSettings(newSettings) {
if (newSettings.longMonth) {
monthStyle = "long";
} else {
monthStyle = "2-digit";
}
if (newSettings.showTime) {
timeStyle = "2-digit";
} else {
timeStyle = "hidden";
}
if (newSettings.weekDay) {
weekDayStyle = "long";
} else {
weekDayStyle = "hidden";
}
if (newSettings.englishNumbers) {
numbersStyle = "latn";
} else {
numbersStyle = "arabext";
}
for (let win of Services.wm.getEnumerator("mail:3pane")) {
win.MahourDate.MahourDateHeaderView.destroy();
win.MahourDate.MahourDateHeaderView.init(win);
}
},
},
};
}

close() {
ExtensionSupport.unregisterWindowListener(EXTENSION_NAME);
for (let win of Services.wm.getEnumerator("mail:3pane")) {
unpaint(win);
}
}
};

function paint(win) {
win.MahourDate = {};
Services.scriptloader.loadSubScript(
extension.getURL("content/customcol.js"),
win.MahourDate
);
win.MahourDate.MahourDateHeaderView.init(win);
}

function unpaint(win) {
win.MahourDate.MahourDateHeaderView.destroy();
delete win.MahourDate;
}

To be able to call the addWindowListener and changeSettings functions, they are defined using the pattern specified in the Mozilla guide inside getAPI(context).

The addWindowListener function calls paint and unpaint functions via the onLoadWindow and onUnloadWindow events respectively. The paint function calls the content/customcol.js script to generate the column content and display it. The unpaint function stops the column display and clears its content.

The default values for the configurable variables are defined here. When they are changed in the extension’s options page, the changeSetting function is called via background.js, which rebuilds and displays the date column.

The content/customcol.js file:

// This Source Code Form is subject to the terms of the
// GNU General Public License, version 3.0.
var { AppConstants } = ChromeUtils.import(
  "resource://gre/modules/AppConstants.jsm"
);
var { Services } = ChromeUtils.import("resource://gre/modules/Services.jsm");

const jalaliDateColumnHandler = {
init(win) {
this.win = win;
},
getCellText(row, col) {
var date = new Date(this.getJalaliDate(this.win.gDBView.getMsgHdrAt(row)));
var currentDate = new Date();

    //fixed options
    var yearStyle = "2-digit";
    var dayStyle = "2-digit";

    var locale = "fa-IR-u-nu-" + numbersStyle + "-ca-persian";

    var year = date.toLocaleString(locale, { year: yearStyle });
    var month = date.toLocaleString(locale, { month: monthStyle });
    var day = date.toLocaleString(locale, { day: dayStyle });
    var weekDay =
      weekDayStyle != "hidden"
        ? date.toLocaleString(locale, { weekday: weekDayStyle })
        : "";
    var time =
      timeStyle != "hidden"
        ? date.toLocaleString(locale, {
            hour: timeStyle,
            minute: timeStyle,
            hour12: false,
          }) + " ،"
        : "";

    //fix for bug that doesn't prepend zero to farsei
    if (time.length != 7 && timeStyle != "hidden") {
      var zero = numbersStyle === "arabext" ? "۰" : "0";
      time = zero + time;
    }
    var isCurrentYear;
    if (currentDate.toLocaleString(locale, { year: yearStyle }) == year) {
      isCurrentYear = true;
    } else {
      isCurrentYear = false;
    }
    var isCurrentDay;
    if (date.toDateString() === currentDate.toDateString()) {
      isCurrentDay = true;
    } else {
      isCurrentDay = false;
    }
    var isYesterday;
    var yesterdayDate = new Date();
    yesterdayDate.setDate(currentDate.getDate() - 1);
    if (date.toDateString() === yesterdayDate.toDateString()) {
      isYesterday = true;
    } else {
      isYesterday = false;
    }

    var placehodler;
    if (monthStyle === "long") {
      placeholder = "TT \u202BWD DD MM YY\u202C";
    } else {
      placeholder = "TT YY/MM/DD WD";
    }

    //remove year if it's current year
    if (isCurrentYear) {
      placeholder = placeholder.replace(/YY./, "");
    }
    //only show time if it's current day or yesterday
    if (isCurrentDay) {
      placeholder = "TT Today";
    } else if (isYesterday) {
      placeholder = "TT Yesterday";
    }

    dateString = placeholder
      .replace("YY", year)
      .replace("MM", month)
      .replace("DD", day)
      .replace("WD", weekDay)
      .replace("TT", time);

    return dateString;

},
getSortStringForRow(hdr) {
return this.getJalaliDate(hdr);
},
isString() {
return true;
},
getCellProperties(row, col, props) {},
getRowProperties(row, props) {},
getImageSrc(row, col) {
return null;
},
getSortLongForRow(hdr) {
return 0;
},
getJalaliDate(aHeader) {
return aHeader.date / 1000;
},
};

const columnOverlay = {
init(win) {
this.win = win;
this.addColumns(win);
},

destroy() {
this.destroyColumns();
},

observe(aMsgFolder, aTopic, aData) {
try {
jalaliDateColumnHandler.init(this.win);
this.win.gDBView.addColumnHandler(
"jalaliDateColumn",
jalaliDateColumnHandler
);
} catch (ex) {
console.error(ex);
throw new Error("Cannot add column handler");
}
},

addColumn(win, columnId, columnLabel) {
if (win.document.getElementById(columnId)) return;

    const treeCol = win.document.createXULElement("treecol");
    treeCol.setAttribute("id", columnId);
    treeCol.setAttribute("persist", "hidden ordinal sortDirection width");
    treeCol.setAttribute("flex", "2");
    treeCol.setAttribute("closemenu", "none");
    treeCol.setAttribute("label", columnLabel);
    treeCol.setAttribute("tooltiptext", "Solar Date");

    const threadCols = win.document.getElementById("threadCols");
    threadCols.appendChild(treeCol);

    // Restore persisted attributes.
    let attributes = Services.xulStore.getAttributeEnumerator(
      this.win.document.URL,
      columnId
    );
    for (let attribute of attributes) {
      let value = Services.xulStore.getValue(
        this.win.document.URL,
        columnId,
        attribute
      );
      if (
        attribute != "ordinal" ||
        parseInt(AppConstants.MOZ_APP_VERSION, 10) < 74
      ) {
        treeCol.setAttribute(attribute, value);
      } else {
        treeCol.ordinal = value;
      }
    }

    Services.obs.addObserver(this, "MsgCreateDBView", false);

},

addColumns(win) {
this.addColumn(win, "jalaliDateColumn", "Date");
},

destroyColumn(columnId) {
const treeCol = this.win.document.getElementById(columnId);
if (!treeCol) return;
treeCol.remove();
},

destroyColumns() {
this.destroyColumn("jalaliDateColumn");
Services.obs.removeObserver(this, "MsgCreateDBView");
},
};

var MahourDateHeaderView = {
init(win) {
this.win = win;
columnOverlay.init(win);

    if (
      win.gDBView &&
      win.document.documentElement.getAttribute("windowtype") == "mail:3pane"
    ) {
      Services.obs.notifyObservers(null, "MsgCreateDBView");
    }

},

destroy() {
columnOverlay.destroy();
},
};

The main content of the column is built here and the new column is created.

Background Script

Background scripts, as special pages, have the window property with the Document Object Model property.

Background page properties:

The background page can access WebExtension APIs when the extension has the necessary permissions. It can also send XHR requests based on host permissions (host permissions). For more information, search for Cross-origin access.

The background script does not have direct access to web pages, but it can load content scripts (content scripts) in them and communicate with them using the message-passing API. This approach is used in the Mahour extension.

Note: Background scripts also have limitations to prevent access to unauthorized actions. The inability to call eval() is an example of these limitations. For more information, read Content Security Policy. Also, calling alert(), confirm(), or prompt() is not possible in this page.

Mahour’s background script code:

"use strict";

/*
Default settings. Initialize storage to these values.
*/
var datePrefrences = {
longMonth: true,
showTime: true,
weekDay: false,
englishNumbers: false,
};

/*
Generic error logger.
*/
function onError(e) {
console.error(e);
}

/*
On startup, check whether we have stored settings.
If we don't, then store the default settings.
*/
function checkStoredSettings(storedSettings) {
if (!storedSettings.datePrefrences) {
browser.storage.local.set({ datePrefrences });
}
}

function repaint(newSettings) {
browser.MahourDate.changeSettings(newSettings);
}

const gettingStoredSettings = browser.storage.local.get();
gettingStoredSettings.then(checkStoredSettings, onError);

/* globals browser */
var init = async () => {
browser.MahourDate.addWindowListener("hich");
};

init();

The storage permission is obtained in manifest.json to use browser.storage.local.get() and browser.storage.local.set(). MahourDate is the identifier for this extension’s new API. Using browser.MahourDate.changeSettings(newSettings), one of its functions is called to apply settings.

Update for Thunderbird Version 115 and Later

In Thunderbird version 110, support for Custom Column was discontinued and the Mahour extension stopped working. With the addition of a new API in version 115, it became possible to update the code and support these versions. If you’d like to know more, we discussed it here.

Since I didn’t have enough time to update the code, it took a while, but eventually Mahour supports newer versions up to 128.

Useful Resources:

If you’re not familiar with building extensions and are just starting out, watch this video by Shojaei:

To test extensions on Thunderbird, in the Add-ons Manager page click on Tools for all add-ons ( ). By selecting Debug Add-ons, a new tab opens where you can temporarily install and test extensions. Simply click on Load Temporary Add-on and select the extension’s manifest.json file.

• Extension API development guide
• Browser extensions
• Simple examples