Do you have any actual problems with systemd, or do you just want SysV init scripts to stick around forever?

Maybe systemd isn’t the best, but it’s way better than a bunch of mostly unstructured shell scripts, and more secure (it’s pretty easy to reduce privileges, sandbox the filesystem, restrict syscalls, etc per service just by editing the unit file)

I use chezmoi and chezmoi_modify_manager to keep my dotfiles (including some KDE configs) in a Git repo, and it works well enough.

I like Pushover too. I’ve been using it for over 10 years now.

Maybe they’ve cast Jack Black as Ganondorf lol

Chrome’s had it for five years, and I don’t recall any Webserial-specific vulnerabilities in it (but I could be wrong!)

you need at least two NICs to properly setup a firewall.

I’m not sure I’d recommend it, but two (or more) VLANs on a single NIC would work fine too. This setup is usually referred to as “router on a stick”

I’m not sure about other OSes or Linux distros, but it’s easy to add multiple VLANs on Debian. You load the 8021q kernel module, then add interfaces suffixed with the VLAN ID (e.g. if your NIC is ens3, you’d add ens3.10 to /etc/network/interfaces for VLAN 10). You’d also need to make sure the switch port is configured to allow VLAN10.

Older NICs lead to regular crashes and/or slow network speeds.

but the ones you’re suggesting (I350-T2 and -T4) are 12 years old.

I have some colleagues from Singapore, and from what I’ve heard, they have to do something similar there.

The weird thing to me in the US is that they’re making the enrollment for Selective Service automatic, but they can’t make enrollment for voting automatic. 🤔

not dependent on the server

It doesn’t have to be - a developer could also provide a HTML file that the user can download and open locally.

And to be honest, if someone had to build a user-friendly cross-platform GUI app for connecting to some sort of serial device, they’d probably just end up using web technologies (Electron or Tauri) anyways. May as well avoid the extra overhead of Electron.

You can completely disable the API in Chrome… I assume Firefox will allow this too.

IMO it’s fine since you need to explicitly grant permission for the site to use it, and also explicitly choose a device to allow it to communicate with. You can also configure your browser to always reject requests to use the API, if you never want to use it.

WebSerial is useful for the developer as they can build their webapp once and it’ll work consistently across platforms, and it’s useful for the user since the same interface will work across all OSes.

I prefer it over the other common approach for communicating with serial devices, which is often to only make a Windows app and to have some convoluted setup process involving sketchy-looking drivers, which then breaks when you have different devices that require different versions of the flashing software or drivers.

Why install a native app when a website can do it? It’s very common to use a website to flash ESPHome for example.

My guess is that they’d have a pool of accounts, and cache the most popular songs rather than streaming them from Tidal every time.

Sorry, I didn’t mean to say they’re the same. I meant to say that if all songs on an album are by one artist, the Artist and Album Artist will be identical. This is the case the majority of the time.

The major exceptions are collaborations (like you said), and compilations (which have “Various Artists” as the Album Artist)

We still call them weeds. It’s called a whipper snipper because it snips weeds by whipping them.

Tidal would be seeing their site URL in the referer for the network requests.

Using it through Lidarr just uses the search feature in slskd, so it might not make it much better.

It’s even open-source! Nice site.

it had the ARTIST tag copied to the ALBUMARTIST tag

This isn’t wrong though - it’s a proper use of both tags. I think most of my music has both tags populated.

That site is pulling from Tidal, which is why the tags are good. All the legit streaming sites have well-tagged files.

Yes. The search results and music files are coming directly from Tidal, using someone else’s account. If you look in the network tab in the browser’s dev tools, you’ll see requests to Tidal.

Interesting design, since it’s trivial for Tidal to block something like this - they can see that the requests are coming from that site. I’m surprised they haven’t blocked it.

yt-dlp has a strict policy against cracking DRM

This is how it stays legal in the USA. Bypassing DRM is a DMCA violation (section 1201), but just downloading freely-available content is totally legal.

Its predecessor, youtube-dl, was subject to DMCA takedowns from the RIAA, and they had to get the EFF to help. yt-dlp doesn’t want to experience the same issues.

It’s also supported by Prowlarr if you want to automate downloads using Lidarr.

Having said that, note that many uploads on rutracker are raw CD dumps (ISO file, plus a CUE file specifying when the tracks start and end) which Lidarr doesn’t support directly, so you’ll have to manually convert to FLAC and split it yourself. Once you do that, you can manually import the files into Lidarr and it’ll tag and arrange the files for you.