Not ridiculous at all, depends on your threat model. It's perfect for me.

So why do we need e2e encrypted messaging? Why do we worry about the "secure" part of messaging? One of the main reasons for me is, it's hard to trust a centralized authority.

Can you trust your information stored at your messaging app's servers stay secure forever? Can you trust that company to never get compromised? So yeh, they go through all the hassle of making things "secure" but attach everything to something(phone number) most of us can't get without revealing our actual details. So yeh, its ridiculous to go through all this hassle to make things private and secure, but force people to use something that de-anonymizes everything about them.

I don't want them to have access to my messages and make a profile out of me, even less of billions of people. That's why I also care about metadata: social graph is valuable for profiling.

I don't write anything that would be interesting to the NSA, I'm not a target for them, so I don't care if they see I use Signal.

BTW, say you have an anonymous Telegram username, I'm pretty sure it's trivial to find your name from the content of your messages. Or just from the metadata.

Is there another technique for effortlessly bootstrapping a contact list? Finding someone requires some kind of lookup/search. Using PII makes a lot of sense because that's how people identify each other. Anything out-of-band is quite complex.

If I told my mom to install some app and we'd connect to each other using some random IDs, we'd probably still be using SMS.

>If I told my mom to install some app and we'd connect to each other using some random IDs, we'd probably still be using SMS

This is surprising to me, since it was my mother's generation that got me into Skype (where you have to manually add the other person by their username), almost twenty years ago.

A Signal support claims "The Signal service does not have any knowledge of your contacts."

Can someone explain how this is possible? How does my phone know when to notify me that a friend has joined signal?

Edit: apparently it's "private contact discovery", I need to give this a proper read later.

https://signal.org/blog/private-contact-discovery/

Yes, their blog posts are really good. Read about the sealed sender, too! And then realise that even with sealed sender, if you send from your home IP, then you are identifiable. So you should send from a reasonably anonymous IP (w.r.t. your threat model)

I will never get over the fact that the Nintendo Switch operated on random character strings to find friends rather than usernames or something more straightforward for normal users. But the fact that they managed it does mean it's not so large a barrier.

I guess it works because people think of it as if they were phone numbers, plus it removes the barrier of having to choose a username. See also the now defunct ICQ which was fairly popular even outside of tech circles.

Repeat after me:

Security and privacy are not the same thing.

Among many other differences: security is a binary (can an attacker gain access to information) but privacy is a spectrum (what kinds of information do I want to leak).

> security is a binary (can an attacker gain access to information)

Only if you're looking through a very narrow lens.

There are a range of potential attackers out there with different motivations and resources available to them. The question you have to ask (for each attacker) is whether the value of the thing you are trying to secure is worth more to them than the cost it would take them to break your security.

In aggregate, that looks a lot like a spectrum of security, with different people being secure against more or fewer different attackers.

"The thing" you are trying to secure is either secure or it is not. That is what I meant by "binary".

"The thing" you are trying to keep private is either private or it is not, too.

Maybe you're saying that some information is more sensitive than others, which is true, but some security failures are worse than others (e.g. denial of service vs. remote code execution, or ability to forge messages vs. ability to decrypt messages).