I use something like this:

#!/bin/bash

set -euo pipefail
URLS=(
    'https://addons.mozilla.org/en-US/firefox/addon/ublock-origin/'
    'https://addons.mozilla.org/en-US/firefox/addon/privacy-badger17/'
    'https://addons.mozilla.org/en-US/firefox/addon/passff/'
    'https://addons.mozilla.org/en-US/firefox/addon/copy-plaintext/'
    'https://addons.mozilla.org/en-US/firefox/addon/duckduckgo-for-firefox/'
    'https://addons.mozilla.org/en-US/firefox/addon/user-agent-string-switcher/'
    'https://addons.mozilla.org/en-US/firefox/addon/clearurls/'
    'https://addons.mozilla.org/en-US/firefox/addon/temporary-containers/'
    'https://addons.mozilla.org/en-US/firefox/addon/consent-o-matic/'
)

DOWNLOAD_BASE_URL='https://addons.mozilla.org/firefox/downloads/latest'
_="${FIREFOX:=firefox}"
_="${DST:=/usr/lib/$FIREFOX/browser/extensions}"

if [ $UID -eq 0 ]; then
    SUDO=
else
    SUDO=sudo
fi


download_links=()
for url in "${URLS[@]}"; do
    package_name="$(sed 's_/$__' <<< "$url" | awk -F/ '{print $NF}')"
    download_links+=("$DOWNLOAD_BASE_URL/$package_name/addon-$package_name.xpi")
done

workdir="$(mktemp --directory)"
cd "$workdir"
for url in "${download_links[@]}"; do
    curl -OL "$url"
done

for ext in *.xpi; do
    ext_id="$(unzip -p "$ext" 'manifest.json' | jq -r '(if .browser_specific_settings then .browser_specific_settings else .applications end).gecko.id')"
    target="$DST/$ext_id.xpi"
    echo "$ext -> $target"
    $SUDO install -Dm644 "$ext" "$target"
done

That doesn’t handle the extension config though.

Google: Laughs in “Everybody else you communicate with who has that shit enabled”

Did you waive your rights to the code anywhere? If not then it’s still your code and they used it without obtaining the rights to using it - depending on your jurisdiction ofc.

Hmm. The first section about cloud service providers is a bit weird to me. There are providers which “keep my best interests in mind” as part of their business model, backblaze would be one. Their whole idea is to provide a good backup services. Encrypting my data before transit also doesn’t make me worried that it will be accessed by them or any of their employees because they will only get some garbled mess.

Compare that to google, another cloud service provider. Their business model is to make money by selling me ads (foremost), they do that by gathering as much data as possible. Here all my answers would be negative.

This puts me in an awkward spot where I nearly every time answer with “Neither agree nor disagree”, because there is more to it and not because I don’t have an opinion.

The whole deployment is done via ansible, so the ansible source is my documentation.

I’m using Concourse CI

I’m still of the opinion, that your GUI sucks if it needs documentation.

Just don’t bring any device to a protest if you consider bringing a burner device.

Something like that should do it:

i = ~((~i + 1) + ~0) + 1

Not yet, though thats a feature worth looking at. I’m thinking that it should be collections instead of playlists. If you add 3 shows to a playlist only the episodes will appear there, while the collection will only show the tv show (or season, whatever you added).

It just manages a native jellyfin playlist, so that should work just fine.

Yes, but it’s incompatible with the way I handle access control. I think I did it with Remote User authentication, which breaks all the login mechanisms of diverse apps, even though it’s officially supported by the projects. That’s why I only choose projects where the frontend is a PWA or they support oidc.

So I just installed the PWA, which works great.

I didn’t read the whole article, just a cursory glance really, but it seems like that is the exact other way around that I would want it.

I’m thinking of scanning a paper bill with my phone, extracting the text and matching parts of the text to firefly fields, like transaction description, source account, destination account, amount and maybe categories/tags.

I have Firefly III and am really quiet happy with it. I might write a companion program to scan bill though, since doing everything by hand is rather time consuming.

There are minecraft reverse proxies, so, yes, a http proxy will not work, but the general idea is still viable and doable with very little effort.

Set up a few domains all resolving to one IP. Run itzg/minecraft-router and use that to proxy the traffic to different servers based on the domain.

Also, they don’t even need a reverse proxy, but just resolve the domain name to the IP (in the simple case of one domain name per I0). That can be accomplished by hosting their own dns server, editing the hosts file or just pointing a public dns record at the private ip address, which will only work in their network,l.