Hacker News Reader

Comment by rfoo | Hacker News Reader

> Sometimes I really wonder the motives and identities behind the people causing such massive and unnecessary drama

In this case, it's pretty boring. They are just a group of "your average power users" or "wannabe programmers" in their highschool or junior years who happened to be born in China so had some exposure to anti-censorship. Being in their overconfident period of life, they pass by various myth they don't really understand as truth. The community is quite toxic but usually they don't cause trouble outside of their own circle, but when it happens, I don't know how to deal with them either.

They also misuses words like "vulnerability" or "responsible disclosure" because some of them have read a lot of news about security research, thought it is extremely cool but have no idea what it actually means.

View on HN

I've seen a growing number of anti-Signal posts and activism lately, mostly surrounding the (well-documented) design tradeoffs that Signal makes for usability and privacy (such as opting to use phone numbers for usernames, to avoid having to store contact lists/social graphs on Signal servers), or their famous decision not to federate/interoperate.

Perhaps it's just criticism growing in lockstep with Signal's overall growth and notoriety, and there aren't any concerted efforts to discredit Signal and sow doubt about using it because it's harder for the intelligence agencies to surveil. I'd like to live in that world.

I'm not sure that I do.

Did you know Signal, like Tor, was financed by an offshoot of the CIA?

https://www.opentech.fund/results/supported-projects/open-wh...

https://pando.com/2015/03/01/internet-privacy-funded-by-spoo...

Now if I were an Iranian dissident, I would be reasonably confident Signal is designed to withstand the Iranian regime’s interception efforts (but not necessarily traffic analysis). If I were someone on the US government’s shitlist like Edward Snowden or Julian Assange, my calculus would be entirely different. (Yes, I know Snowden has endorsed Signal)

Except Snowden said he uses the app everyday and is still alive. OTF money doesn't come with strings attached, if you can _prove_ otherwise, you might actually have something valuable to contribute to this conversation, but now it's just words that should be dismissed as conspiratorial.

One major reason for the push back against Signal promotion is that it does not represent any sort of federated protocol. It is a complete silo. So if it did become popular it would eventually be a serious problem and would need to be fought against.

That isn't a reason to push against Signal, it is a reason to push for a federated solution.

Not your point, but federation is often heralded as some must-have feature for communication around here, but we already have that: XMPP does that, even with encryption. Or Email with PGP. Or even the old Textsecure code, as Signal/Textsecure started out federated, but, surprise, that didn't help adoption.

> It is a complete silo. So if it did become popular it would eventually be a serious problem and would need to be fought against.

This is an explicit design decision. It used to federate, and they found that to be terrible, so they stopped, and now it's better.

Maybe you should find something else to fight against.

https://signal.org/blog/the-ecosystem-is-moving/

Worth noting: Google Talk used to federate, via XMPP. They found that almost nobody actually used the federation functionality, and something like 99%+ of the federation traffic was inbound spam.

Federated protocols aren't very good, and don't evolve.

> Google Talk found that almost nobody actually used the federation functionality,

Compared to the number of GTalk users, a bunch of self-hosted users probably didn't count for much.

> and something like 99%+ of the federation traffic was inbound spam

I heard this too, not sure if it was the only reason.

Too bad Google doesn't know how to manage spam. Good thing they don't have any other federated communication products. /s

> Federated protocols aren't very good, and don't evolve.

What makes you say this? XMPP has come a long way since 1999 and is still evolving to this day. Even email is slowly evolving. Is Matrix not evolving? HTTP isn't really federated, but Moxie also mentions the web being stuck on HTTP/1.1, because HTTP/2 and HTTP/3 don't exist.

>So if it did become popular it would eventually be a serious problem and would need to be fought against.

It already is.

>use phone numbers for usernames, to avoid having to store contact lists/social graphs on Signal servers

Why can't the user just be expected to deal with their contact list? The phone company doesn't store your address book either.

This is a terrible excuse. They require the phone number as an anti-spam/moderation measure and hide behind privacy.

> Why can't the user just be expected to deal with their contact list?

They are; they let the user maintain their own list of phone numbers for their contacts, precisely like the phone company in your example.

> They require the phone number as an anti-spam/moderation measure and hide behind privacy.

This is (inaccurate) speculation from ignorance. Signal, unlike almost every other phone-number-using-service on the planet, does not block burner/disposable/voip numbers from being used with the service.

Then why not let users sign up with a random unique ID number that they can share with their contacts?

Usernames are coming.

Reply on news.ycombinator.com