Linux config files and scripts
The main branch is desktop
I'm using a slightly modified version of the method described by StreakyCobra and others in this Hacker News thread.
curl -sL https://raw.githubusercontent.com/etacarinaea/dotfiles/desktop/dotfiles-install.sh | bash
..or if you want to do it manually you can also enter the following commands in a shell:
git clone --mirror git@github.com:etacarinaea/dotfiles.git $HOME/.dotfiles
git --git-dir=$HOME/.dotfiles/ --work-tree=$HOME config \
--local status.showUntrackedFiles no
alias config='/usr/bin/git --git-dir=$HOME/dotfiles/ --work-tree=$HOME'
Uses MathJax for LaTeX and Marked for markdown support.
I adapted my configuration from Niklas Korz' version here, which is based on this reddit post.
Additionally, I'm using vim to write cards, because it's a lot easier and more efficient to use than Anki's integrated text editor.
.anki/
contains "config" files (front/back templates and styles) for Anki cards.vim/ftdetect/ankicard.vim
adds a new filetype to vim for anki cards.vim/snips/ankicard.snippets
uses the aforementioned filetype to define snippets for anki, extending the already existing tex snippets
Rebinds caps lock to Super_L
Renames the current i3 workspace while preserving the numbering. Entering nothing removes the name again leaving only the number.
Example screenshot:
Gets called by dunst
when Spotify sends a new notification (see
.config/dunst/dunstrc
). Writes the current title and artist from the
notification into /tmp/spotifylog
which then gets read by i3status (see
.config/i3status/config
).
Disables the function keys so etc/acpi
can handle them instead.
Handles function keys. Keys implemented:
- mute
- volume down
- volume up
- brightness down
- brightness up
Volume is only implemented for pulseaudio
.
All keys also send a notification using libnotify
.
Fn+Esc
(FnLk) and Fn+Space
(keyboard backlight) worked out of the box for
me.
systemd service for locking the screen with i3lock
It also uses ImageMagick to capture the screen and then display a blurred version with some text overlaid as the lock screen's background.
The normal -blur
function is pretty slow for high sigma (and -gaussian-blur
is even slower), so depending on the system it could take a while for convert
to run, which delays locking.
Scaling the image by some factor A < 1 and then scaling it back up by A's multiplicative inverse with a Gaussian filter with some sigma, gives more or less the same blur as applying a Gaussian blur with that sigma times the multiplicative inverse of A without scaling, and is a lot faster.
So for example if you'd usually need a sigma of 40, and you scale the image by a factor of 0.1, you end up only having to use a sigma of 4 to get more or less the same visual effect.
Obviously if the scaling factor is too small you'll end up loosing too much information and the end result will just look like a gradient or a solid colour.