tmpod

Mullvad Browser isn't bullet proof, it will not prevent fingerprinting entirely, though it makes it less reliable, especially if it isn't sophisticated.

Finally! I had tried using the clunky torsocks not long ago and wondered why there was no namespace based solution yet. Glad to see this getting released, it will help many people. Tor ❤️

This is quite misleading and frankly low effort. Besides the readability issues, the chart makes a clear distinction between Proton Pass and Bitwarden when it comes to privacy, citing their privacy policy.

As it happens, however, Proton's server code is closed, unaudited[^1] and not distributed, and the apps (web, Android and iOS) do not support setting different homeservers. This effectively means you cannot self-host your password manager and must be "locked" to Proton for what I consider to be one of the most fundamental and important pieces of technology a person can use.

Bitwarden, however, has opened their official C# server, their internal Rust SDK and the apps themselves too. Furthermore, they have several guides on how to self-host your own personal server, and have implemented settings in their apps to change the homeserver. There's even an unofficial server, vaultwarden that is even better tailored for small, personal deployments.

All this to say: the fact they may collect some usage data on their website is very insignificant for their offering, in my opinion. The real value is in providing a secure vault that only the user can manage. If you need better privacy and/or anonymity, you should use tools specialized for that anyway, instead of blindly trusting a third-party's Privacy Policy, no matter who they are. But then again, it's the old game of threat models.

Ultimately, Bitwarden inspires more confidence than Proton, by giving those you can and want the ability to truly own their secrets.

[^1]: As far as I'm aware, there's only this audit by Cure53, in which they performed a white-box pen test on the API, with only its documentation provided, no code whatsoever. These audits are important from a cybersecurity point of view, but security is not the same as privacy and should not be taken as such.

Adding onto what's already on the thread, you can try look at the newer Element Call, which is an implementation of Matrix's native calls.
I've been using it a bit recently, since Jitsi seems to have stopped working reliably for me (to be frank, I've not put much effort into debugging it yet). It works well, but it's still early stage, lacking some features Jitsi has. If that one works for you, I recommend you stick to it.

Yeah withdraw cash from an ATM and use it. The system sucks, but it's not trivial to change for a myriad of reasons.

There's no real way to do it. Unless you know someone who can trade you XMR<->cash and you somehow convince your employer to (break laws and) pay you in those forms, you can't avoid it. At some point, you'll have to get money on a real bank account, which requires real information to open.

As far as I know, modern cards don't just send your CC info to terminals, they do some form of a cryptographic handshake (probably a pubkey signature or similar) which gets confirmed by your bank. I believe Caveman was talking more about online shopping, where you have to enter your card number, expiration date, CVC and often your name too.

That's why I love virtual card systems like MB NET. You just generate a random virtual card for every purchase (or a recurring one for each subscription vendor, for example) and move on. Your bank still knows what you're doing, of course, but vendors can't correlate anything. Preventing your bank from knowing where you're spending your money is much harder, for very practical reasons: fraud detection. The only real way is to use a secure crypto coin like Monero, but very few places accept it and you still have to deal with volatility.

Ah right, Molly. Have yet to tried it, but looks interesting.

I think I'm too afraid of moving my main stuff to Molly, lest I lose something :P But the UnifiedPush and multiple mobile clients is enticing.

This is a good suggestion. Docker is more mature and has more resources, so it's better to learn the ins and outs of containers. After getting comfortable with it, you can move to Podman and have a much better time tackling its peculiarities regarding permissions and rootless.

I used Docker for years and only recently decided to give Podman a try, porting my Lemmy instance to it.

Yeah that's a bummer. Signal has multi device support but only for desktop and iPad (yeah, not Android tablets), but you always need to have a master phone device.

It's been an issue for so long, but this is Signal, they do whatever the f they want.

encrypted email

Besides being a form of messaging (so the text somewhat contradicts itself), typical email is a deeply insecure protocol.
In my opinion, it's probably impossible to secure without making a new protocol or making such drastic changes that it might as well be considered one.

Here are some key concerns regarding the usual PGP-powered encrypted email:

• Email, at a simple level, works much akin to physical email — there's an "envelope" containing important info regarding the communicating parties, which can't be encrypted, otherwise the mailing servers wouldn't know where to forward the messages. This essentially leaks a lot of metadata that can be almost as valuable as the message body itself.
• There's no forward secrecy — one of the best cryptography features that has become pretty much a commodity in modern systems is forward secrecy, which prevents attackers from decrypting older messages after gaining access to one of the keys.
• While not an issue with the protocol itself, it's the sad reality and we need to consider — most people use GMail, Outlook and the like, which ultimately need to read your emails in plaintext, for better or worse reasons (search is incredibly useful, but some big players don't stop there of course :p).
• Another thing is the fact that it's incredibly easy to have an imbalance of encryption, i.e. someone is encrypting their messages, but others aren't. With the very popular email culture of quoting (be it top or bottom posting), an unencrypted party in the the conversation can leak important information.
• PGP is... peculiar, so to speak. I has a lot of issues, mostly stemming from its age (which could also be a source of robustness and security, due to being very battle-tested, but I don't think that's quite the case with PGP/GPG), tries to do too much and typically has a clunky UI, which impedes wider and proper adoption by less technically people.

This isn't to say people should definitely stop using and promoting encrypted email, since it can be useful.
It's just it gives, more often than not, a false sense of security and can lead less proficient users to send sensitive data through this medium which isn't nearly secure enough for such use cases. Preferably, people with such threat models should opt for better alternatives, most suggested in that article (such as, but definitely not limited to, Signal, SimpleX, Matrix+Olm, XMPP+OTR/OMEMO, sharing files via MagicWormhole, encrypting with tools like age).

On a slightly tangential note, I think someone should make a Matrix client with an email client interface. I started working on a new traditional chat client (completely nonfunctional still, very much in-dev), but I've been honestly thinking more and more about making one looking like an e-mail client, where there isn't much focus on instant room-based chats, but rather on longer-lived 1-to-1 and list-like exchange of messages.