I will second the suggestion for a dash camera, they can record audio automatically without need for user intervention.

Another option is to just use a voice assistant, I usually quickly ask my watch to start a voice recording beforehand to ensure there is a record of the interaction.

If you don’t mind me asking, what music client is pictured?

I think there’s an element of prestige people are missing. At least in my country there were online options prior to the pandemic even, they however lacked the prestige / name recognition that other institutions had. Keeping mandatory in-person classes is another way to maintain this prestige, a differentiating factor, from the other institutions.

I also have to agree with most of the comments here. From an instructional point of view online classes are lacking, they can be less engaging, and pedagogically neutered. And in fields with lots of laboratory work, it’s frankly impossible to get rid of at least part of the in-person educational component. Even for the humanities, having access to a large on-campus library of scholarly resources is integral to research.

In my personal experience I’ve been quite grateful to have access to a large archival collection, items that could not be shipped to remote students because they are too old to leave a temperature & humidity controlled environment. An online experience would prevent someone like me from doing some manuscript / original publication related research.

Now, I do think online options are helpful. ESPECIALLY for summer classes, where students may wish to retake a class while also moving away for summer work. But I do not think they should become the default, they should be an option where possible, but not the new normal.

Normally I would say community forks have the power to continue the project. However, in this case I think chrome / safari would eventually add enough new features that Firefox forks can’t add quick enough. Mozilla at least has some power in pushing the direction of web-standards, which these forks would lack, as well as the larger development team and some corporate usage of the browser which Mozilla has. I also don’t see the smaller development community keeping up with security issues found in the browser, particularly pertinent for corporate marketshare and individuals with a stricter threat model (journalists, dissidents, etc.)

The only other factor, is whether Firefox dissapearing would officially create impetus for an anti-trust case against Google. I doubt so under the current American presidency, but I could see the EU being concerned (even if they lack the power the US has to force the company to split). If something were to happen here there would be substantial change in the browser market, but I wouldn’t be too hopeful of this happening.

One of the advantages of Relay is that it is agnostic of your email provider, making it easier to switch providers without having to change the email on every account that has an alias.

Considering this, I’d be tempted to go with Addy.io instead of ProtonMail / SimpleLogin (subsidiary of Proton AG).

If you’re concerned with having to trust a third-party to process your emails however, Proton may be the better option with built-in aliasing. Mailbox.org is another option recommended by privacy guides with built-in aliasing.

If you’re concerned with Mozilla’s TOS change however, you may also be concerned with the Proton CEO implicitly supporting the current Trump presidency, believing that the Republicans will do a better job reigning big tech in (While I’ll agree that the democrats are not anti-corp, that died with Bernie, I think it’s foolish to believe the republicans will be better). They also pulled their entire media presence on Mastodon, and recently integrated Zoom despite explicitly stating that it has privacy issues in their blog.

I think some people are being a bit extreme in their characterization of Proton AG right now, but it definitely feels like they’re making some peculiar choices when looking at their guiding mission of privacy / security.

What are your goals?

I would say it’s really a combination of the instances policies and their jurisdiction, and in terms of jurisdiction it also depends on where you live (e.g. you may have more protections under law if the instance is hosted in your country)

There’s also nothing stopping you from using multiple instances — siloing your interaction in different types of communities in different accounts on different instances. This may be useful if part of your privacy concerns are having all of your post / comment data on one account on one instance.

Edit: You can also use an email aliasing service to avoid even giving your email out. There are aliasing services such as Addy.io, Simplelogin (subsidiary of Proton AG), Firefox Relay (Mozilla), as well as some email providers which provide (iCloud, Proton, Mailbox.org to name a few)

If you don’t mind me asking, what makes gnome more adaptable in terms of functionality than KDE?

For iOS devices the most up to date client is “Strongbox”. I don’t think it is FOSS, but is compliant with the standard. It’s sadly a freemium app, but is quite well made in my testing. It cannot sync with syncthing, but does support several cloud services, its own service (which uses iCloud), and local file transfer over LAN. They also have a version of the app with all network connectivity removed for security (if you prefer)

This is perhaps overkill, but you can also encrypt the contents of your online cloud storage with CryFS / Cryptomater. This is particularly useful if you wish to store sensitive documents (healthcare, finances etc) in a cloud environment in case of catastrophic destruction of property (destroying computers / on site backups of data).

In this case you can also backup your keepass file in this encrypted virtual storage medium, on top of the prexisting encryption of the database itself.

My personal choice right now is KeePassXC (PC) / KeePassDX (Android) + Syncthing And Aegis (Android) for 2FA codes, with a yubikey for services that support FIDO keys.

Overall I like this setup because it’s decentralized and does not rely on a third party server structure. The only “weak” point would be the Syncthing relay servers or the Tailscale VPN that I use, but this goes back to ensuring encryption of the database is adequate with a long password, and using an open source synchronization protocol that ideally has been vetted by a trusted third party (or yourself if you’re capable)

I used to use Bitwarden, and I highly recommend it. I really appreciated it’s ability to integrate with email aliasing solutions to generate new aliases from within the bitwarden UI itself. However, my main reasons for switching were the following

• I don’t have the money to pay for it (uni student)
• I prefer a more self-hosted approach (I will consider using vaultwarden in the future when I have more money)
• I wanted to move away from using a browser extension for password management on desktop. KeePass’ auto type feature is really good, and a more secure input method than a browser extension autofill.

The only additional advice I have for both recommendations is that I do not think it advisable to add Totp 2fa information to your password manager even if it supports it. I feel like this should be separate, on a single device, and backed up in ~2 locations (one preferably off site). This is really to avoid problems if a device is compromised and if your password manager is compromised, but this is definitely in the more unlikely category I feel.

My only major issues with keepass are the potential for sync conflicts and the some feature differences between platforms. A centralized server config like vault/bitwarden prevents the sync conflict issues, at the cost of having one point of failure. The feature differences problem isn’t too great, but autotype doesn’t work on Linux if you install with flatpak, and you can’t prevent screen capture of the app on Linux (only on Android and Windows from my understanding)

Edit: I also tried gopass, it’s really fun to have an entire CLI based password manager, but frankly the state of mobile companion apps are appalling. The Android option only is good if you use a dev version, and the iOS one I thought was just ok. I also dislike the metadata leaking that is inherent to the format, and that PGP is the main form of encryption for the time being (some clients were looking at using AGE at some point). Overall it’s a cool but flawed concept, and I feel my other two recommendations are superior.

I think it’s important to see these types of efforts, while I’ll never go out and buy a MacBook the effort isn’t wasted since it gives current users more freedom and future people buying used laptops more options for Linux compatible hardware.

Without a project like this, that hardware will end up being e-waste a lot sooner than it should be, when Apple drops support. At least to me I see an ethical and moral imperative for projects like this, but I also understand people’s grievances with Apple.

I think you summed up my thoughts on the matter much better than I could have. In particular, the “digital” / “corporate” right to be forgotten is distinct and much more specific in its scope than a broader right, and is a rather important consumer protection in my opinion.

I have a few shortlisted

My parent’s 2010 Ram 1500, the interior is rather comfy but the reliability is just not there. At 100k km the engine blew up, apparently this is still an issue with the current ones as the 5.7L V8 still has the same flaw allowing for some components to drop into the cylinder. There’s also been random electrical components that have died relatively fast, and whatever metal was used rusted exponentially even with rust proofing being applied twice a year. It had more rust than their 2011 Toyota Highlander that had greater than 300k km

I also just hated when I had to drive it downtown, but I can’t exactly blame the vehicle for that.

2011 Toyota highlander, it went through 3 transmissions, 5 rear wiper motors, and it was about to go on to its 4th transmission when they sold it. The 3rd one didn’t even last much more than a year.

2006 Rav 4 (V6), this car also went through 2 transmissions, and then had to have the entire steering column replaced by year 2

~2016 Ford Fusion, this was a rental car for when my Civic was being repaired after an accident and my god was it awful. It handled like a massive boat despite being a medium sized car and the transmission felt significantly less responsive than even the CVT in the honda. The seats also sucked but i think that was how the rental company cleaned them, they made this awful noise every time you sat in them and looked and felt like a “casting couch” with several generations of children dried up in them…

Honorary Mention: my friends Nissan Versa, seemingly unreliable and falling apart but it refuses to ever give up. That thing will survive nuclear winter, and will remind you with every pothole that its existence is torture.

The plugin that brings the “starter” / “welcome” screen when nvim is called without a file is mini.starter, a lua module of the mini plugin. My primary use case for neovim is closer to a feature complete text editor rather than a full fledged IDE, although there definitely is some overlap in my setup.

My set of plugins are roughly as follows

• vim-plug, I will likely replace this one with packer at some point
• goyo.vim and limelight.vim for distraction free viewing and editing
• nnn.nvim to integrate the nnn file manager into neovim
• mini.nvim according to the Github, “Library of 35+ independent Lua modules improving overall Neovim (version 0.7 and higher) experience with minimal effort. They all share same configuration approaches and general design principles.”
  • mini.surround feature rich surround actions
  • mini.statusline a very simple no-frills statusline
  • mini.starter aformentioned start screen
  • mini.pairs inserts the paired character, e.g typing ( will automatically place ) behind the cursors
  • mini.move move selections
  • mini.map has a little map of the file similar to VScode among many other IDEs & text editors
• barbar.nvim Tabbar plugin
• a whole bunch of LSP / autocomplete engines / snippets / git commit messages & signs
• nvim-treesitter for syntax highlighting

And the remaining things in my init.lua file are just keybindings, setting up the plugins, and disabling the swapfile etc. when editing my password secrets in gopass among other ‘secret’ files

It definitely is rather reminiscent of older Windows versions with the seperate application launchers, fully expanded task bar entries that show the name of an app that are ungrouped (until necessary). And the widgets are very reminiscent of Rainmeter.

But I also bring some things from macOS that I enjoyed such as the global menu on the top (sadly Firefox flatpak does not support), virtual desktops with the pager widget on the bottom, and I use Krunner a lot (plasma’s equivalent to macOS “Spotlight”)

I hope your switch to Linux goes well if / when you switch!

I understand why they wouldn’t want to suddenly change the branding of existing projects though.

I’m not sure if I agree, I feel like the long term damage of keeping the names is greater than changing them now to Fedora Plasma Atomic (Formerly Kinoite) / Fedora Atomic Workstation (Formerly Silverblue). Leaving them as is, is just going to create more confusion in the future to new users who won’t immediately understand why the naming convention is different for the other spins and will create more confusion for documentation / support threads online.

I feel that I am 50:50 on it, immutable at least conveyed more information about what it is while Atomic feels a lot more “buzz-word-y” and does not convey as well what it means. Regardless, I’d say the bigger issue is keeping the old Silverblue & Kinoite names, they really should change them even if it means having a ~2 year period of having “Formerly Silverblue / Kinoite”.

Thank you for the very thorough reply! For god knows what reason I get this error: error: app/org.mozilla.firefox/x86_64/stable not installed when running the xdg-open firefox-reader command, yet manually running flatpak run --user org.mozilla.firefox about:reader?url=https://example.com works just fine. I’ll have to troubleshoot it when I have a bit more time ;p

Thanks again for your very thorough write up and the linked articles. Have a good day :)

Update: It seems like on my system, the --user flag was the issue, removing it made the script function. I am using Fedora Kinoite (Immutable version of KDE Plasma), so perhaps it is just a difference in how flatpak is configured between distros? I’ll have to read into it more later.

I’ll keep my answer focused on KDE Connect as I no longer use a TWM. You can most definitely use KDE Connect in non-Plasma environments. For non-Plasma (and non-Gnome ^* ) environments you can just install the kdeconnectd package. Then, to start the KDE Connect daemon manually, execute /usr/lib/kdeconnectd. You can schedule this to autostart as a systemd unit, or in the config for your TWM (I know in sway/i3 you could start it, I’m assuming it is similar for many other options)

If you use a firewall, you need to open UDP and TCP ports 1714 through 1764. If you use firewalld specifically, there’s an option to enable KDE Connect rather than manually specifying it. This also let’s you have it only work on private networks and not public if you so chose.

See Arch wiki for more details

*For gnome I would recommend using gs-connect even if you have a tiling extension

£ KDE-Connect: does that work on TWMs? Is there a good implementation? Can I use GSConnect elsewhere too?

I’d much rather use a separate Firefox (now Mozilla I think) account for my professional work. I also would prefer having separate extensions, notably Zotero connector is kind of useless for my personal browsing