Live data from Hacker News

Looking back at how Signal works

signal.org

241–250 of 301 posts

Re: Looking back at how Signal works

#241

Signal is a walled garden. They refuse to allow federation and even prohibit any modified client to use their servers. It's the least open "open source" model, and once (if) they gain significant market share they can easily close down the app and lock-in the users. Please use and spread federated alternatives. Donate and contribute.

What alternative would you recommend?

Started using Jami (GNU Ring). No central servers, no phone numbers.

Re: Looking back at how Signal works

#242
post #199

I love Signal and use it as much as I can, but I'm thinking of switching to Matrix solely because the desktop client is pretty bad. It won't show me messages until it syncs everything (so I can't even see old messages while things sync), and, what's worse, it skips messages, and multi-device just doesn't work. My laptop just shows "Message could not be decrypted" until I delete everything and reset. I'm not sure why…

Riot and almost all of the other clients I tried all phone home to some kind of metadata/identity or push server that is run by some private company I’ve never heard of (which isn’t part of Matrix/Riot). Check out its network traffic sometime.

Riot only ever talks to identity servers (which simply maintain a directory of matrix users by phone number or email address) if you actually try to look up a user by one of those identifiers, and if you actually opt in to it. We used to do it without opt-in when you searched, but this was fixed at https://matrix.org/blog/2019/09/27/privacy-improvements-in-s....

For Push, whatever app you use needs to have a push server that talks through to Apple/Google if you use their push. For Riot, that server is run by the New Vector (vector.im), the outfit which makes Riot.

The network traffic should make this very clear.

Re: Looking back at how Signal works

#243
post #213

Earlier quoted context omitted.

The way you use encryption matters as much as the quality of the crypto used. In Telegrams case, usage is all wrong and not even better than e.g. WhatsApp. Opt-in E2EE is worse than having it be default, and server side encryption with server side keys is bordering on the pointless.

Thank you for reflecting with an opinion. I also believe that opt-in E2EE is worse than having it by default. We have precedent for E2EE chats by default and syncing those E2EE messages across devices using the latest RiotX for Matrix. I'd love to see Telegram adopt that strategy. I use Telegram because it delights its users. When I have tried to bring friends and family to first Riot, then Signal, the experience I e…

On the Matrix side we’re removing the recovery passphrase shortly; it was a misstep.

Re: Looking back at how Signal works

#244

Earlier quoted context omitted.

> Meanwhile, yes, Matrix took years to add encryption, but it works much better than Signal, even with quite a few small bugs. I'm not sure which Matrix client you use, but clients like Riot don't even let you opt out of sending read receipts unless you edit `/etc/riot/config.json` to enable experiments and then go into the settings to disable read receipts. Problems like this (and issues like this [0]) give me the i…

I'm a massive Matrix fan and have high hopes for it but in experiments we've done with activist and journalist partners we've found the Riot.im client often gets a bit complicated for people to use. I think the main issue people have is related to keys. As I techie I love the options but I find many don't like having all the options. Signal of course is a lot easier as it hides many of those issues in the UI/UX.

The root problem is the requirement to verify that you are talking to who you think you are talking to. If you skip the identity verification stuff then you are inherently trusting a 3rd party. So if you are not exchanging your key fingerprints (safety numbers in Signal terms) then you are kidding yourself.

Re: Looking back at how Signal works

#245

Earlier quoted context omitted.

I knew Signal was against federation but I hadn’t realized they had pretty much banned third party clients. That would otherwise have been a really easy win for people that actually care about the system integration, performance, and architecture of desktop clients enough to shun electron “clients.”

The Signal team has always been open about the reason why they reject third-party clients: they claim that XMPP adoption was hindered by the inability of a user’s software to know if the software on the other end supports the same feature set. XMPP had grown into a large set of features that some clients supported and others did not. If Signal introduces a new feature, it knows that all users’ devices will support th…

Ironically OMEMO for XMPP ended up being based on the Signal protocol.

Re: Looking back at how Signal works

#246
post #220

Earlier quoted context omitted.

I'm a massive Matrix fan and have high hopes for it but in experiments we've done with activist and journalist partners we've found the Riot.im client often gets a bit complicated for people to use. I think the main issue people have is related to keys. As I techie I love the options but I find many don't like having all the options. Signal of course is a lot easier as it hides many of those issues in the UI/UX.

Out of curiousity, did you ever try keybase? It always struck me that usability as well as security were their primary focus. And I think they did the whole key management / chain of trust thing really well. Hopefully none of that changes now that they've been acquired by zoom.

Keybase is not end to end encrypted and can't be compared to Riot or Signal.

Keybase claims that it is, but it's a lie that nobody verified: https://security.stackexchange.com/questions/222055/how-can-...

Re: Looking back at how Signal works

#247
post #64

Earlier quoted context omitted.

> I'm thinking of switching to Matrix solely because the desktop client is pretty bad. Oooooh, I've got some bad news for you.

What is it? Riot works beautifully.

Beautifully mediocre, but yeah it mostly works. (Never been able to use Signal so I can't compare, but Riot's encryption was completely broken until a few weeks ago, and now there is only the occasional "hold on, gotta fetch keys for 10 seconds" and slow scrolling up, etc. It can't compete with something like Telegram in terms of UX; it is infinitely better for privacy but it doesn't "work beautifully".)

Re: Looking back at how Signal works

#248
post #69

Earlier quoted context omitted.

The legal entities can move to other jurisdictions, sure, but it doesn't matter because app distribution still occurs primarily through USA-based Google Play and USA-based Apple App Store—both of which can easily geofence apps as they please (or as they're required). This is one of the reasons I've started to appreciate Matrix a lot more lately. https://matrix.org/blog/2020/01/02/on-privacy-versus-freedom

The real solution is switching to open source mobile operating systems.

Android is open source. Alternative app stores exist. Play Services can be replaced.

Re: Looking back at how Signal works

#249

Is it worth trying to move my friends from WhatsApp to Signal? As I understand it, they're both e2e encrypted. I'm also trying to move my chats from SMS and Gchat to something encrypted, but am torn between WhatsApp and Signal. The former has more of a buy-in with my contacts already. I realize WhatsApp is owned by Facebook, but isn't the whole point of e2e encryption that you don't have to trust the intermediate inf…

WhatsApp is opportunistic encryption by default, trusting the server to provide it keys. Pretty much nobody turns on showing when someone's key changes, and if you do and you verified each 's keys, the server can still learn the plaintext of at least one message by changing keys right when you send a message. Which is a very tiny risk, but it's just not what one would expect. It's as if your browser would continue loading a page with an invalid HTTPS certificate and show you a warning afterwards -- and only if you turned on showing that warning in the settings.

Even if you turn it on and you're not afraid of the server swapping the key when you send a message (which is fair enough), do all your contacts also have it turned on? And would they tell you immediately and out of band (otherwise the attacker can suppress or change the message) that they saw your key change?

But an even bigger reason I don't have whatsapp is because of Facebook. Metadata has already been mentioned by others so I won't repeat that.

Re: Looking back at how Signal works

#250
post #217

Earlier quoted context omitted.

I, for one, have a bigger problem with it forcing the use of phone numbers as a sign-in method. They're an arbitrary identifier from a legacy system that there's not really a point in continuing to extend, because if your device is capable of anything more advanced than SMS it's also capable of... well, this. Also KaiOS and the like are making chat feasible even on feature phones. Don't get me wrong, RCS will be a fi…

"even on feature phones"? It was perfectly possible at least back in the early 2000's. Where I'm from, we've gone through a lot of different IMs over the years, including XMPP (which I was a big fan of, but started to despise because it had such terrible support for mobile clients). Many J2ME clients had pretty advanced features like group chats and file transfers.

conversations.im is an excellent android client for xmpp.
Post reply on HN