Agree, more possibility by using more code usually comes with more risks.
I think the vision is different and this has a big impact in everything. I searched but could not find the reference, but I read on simplex chat blog that they are trying to reach a mass adoption of the software broadly in order to improve the privacy of everyone on the planet.
In order to reach this goal, they needs to offer a very good ui and all of the features that are commonly offered by non private messengers.
The quote was something like: adding sticker support to our application seems like a loss of time, but if the missing support of sticker is what prevent everybody to join simplex, then in order to improve privacy globally, we need to add sticker support. Sorry that I could not find a reference to the quote.
So yeah… If you need to configure tor or if you can only send message when both people are online, its not usable by the vast majority of the population thus its not going to won or improve the market. Its fine to develop a tool for the few person with a higher thread model, but with all the difficulty of uses, it will probably not go worldwide.
In another thread about simplex, maqp mapq ? Write a nice message about simplex not advertising their capability accurately. I tend to agree. However he was also requiring a metadata resistant messenger, which might be more on the higher thread level with specific person in mind, and not a software that aims to reach global usage for common folks.
Think you might be misremembering. It isn’t maqp who is holding SimpleX to any unreasonably high standard, it is SimpleX’s marketing that is (at least according to maqp).
Yeah, like I said in the thread, an app that sends a copy of every message to the NSA, that openly tells it sends a copy of every message to the NSA, is a secure app.
It’s the what’s advertised not meeting what’s offered, that’s the problem. I’m not against SimpleX if it accurately conveys its threat model. I’m against anything that doesn’t, and the app is only one of half-a-dozen I’ve criticized.
I also have zero problem changing my opinion. I complained about iMessage for years, then it upgraded from RSA-1280 into the post-quantum thingy with safety numbers and all. Now I think it beats WhatsApp. But since it’s proprietary, it doesn’t go higher than Signal, even if the protocol is apparently slightly better on paper. I used to complain about Matrix and its crappy E2EE across clients, and now I have no trouble recommending it for large institutions who need E2EE.
I have no trouble recommending SimpleX once its walk matches its talk.
I’m not against SimpleX out of principle, I’m for the users out of principle.
SimpleX is cool but did they add the feature to send things like stickers? I dropped it long ago because of that. My friends and family only accepts Signal because it has those friendly things.
SimpleX is one of the strongest options for privacy, Briar also, but I chose Simplex. By proof I mean if somebody has a better app for privacy and to prove why.
It is, but sad at same time. It is hard to find people born after the 2000’s that accepts things like chat communication without those “conveniences”. Anyways, the most secure is probably not the case where my family fits. If I was an activist or journalist I’d be looking at SimpleX and Briar.
Completely agree SimpleX isn’t going to be used by me or my friends any time soon, no matter how much I care about my privacy. That’s the unfortunate reality we live in It’s just so funny to imagine that someone won’t use X tool/service because of Y feature (or lack thereof) despite Y being the least of your worries.
It’s a valid complaint when you take into account social dynamics. Is it a valid security complaint? Definitely not. Good and bad depends on the context. Is it bad security? no. Is it bad for advancing privacy to the masses? yes.
I’m surprised nobody has mentioned the fact that Jack Dorsey is the majority funder of SimpleX. That means he has majority control over it. He co-founded Twitter which started off with a ‘pure’ platform but morphed into working with government banning, censoring and surveilling it’s subscribers to fit a desired narrative/agenda. To me that alone is a major reason to consider SimpleX a major risk.
Nah. No tech company deserves your trust. It doesn’t matter who’s running the show, because (nation state) hackers are going to breach the vendor side at some point anyway, so you assume the worst from the vendor, always. SimpleX, like every app out there, is exactly as secure as the the client let’s it be:
Is it peer-to-peer so there’s no server with access to metadata? No
Does it anonymize connections to server with Tor by default so server can’t tell who the users are by their IP? No, so it can infer who’s talking to who.
Does it feature traffic masking flow confidentiality to hide when, how much and what type of communication takes place? No.
Do we have court evidence to show they have nothing collected about their users? AFAIK no.
So SimpleX is content-private by design, and not metadata-private at all.
Use it under that threat model and it’s fine. Need metadata protection? Look into more secure alternatives like Cwtch, Briar, and Quiet.
Yeah more or less. You can use say Tox over Tor, or you can use messaging tools that operate via p2p model using Onion Services and web clients.
Doesn’t padding count as traffic masking which SimpleX already does?
That’s just basically rounding to nearest block size with say PKCS#7. It has nothing to do with traffic flow confidentiality (should’ve used that term above, I’ll fix it) where you have continuous stream of noise data being transmitted, and into which you inject packets you send.
Sealed sender means the server just strips the sender information when they write it to cache. If the server is not compromised in that malware or malicious developer puts the feature back it makes the cached ciphertexts much less useful to attacker as they can at most infer who is popular. So it’s not nothing in practice; we can see in court documents Signal does indeed not have that data to hand out. But as per the criteria of “is it metadata protection by client”, nope.
As for the proxy, I’m unsure what you refer to, the TLS proxies? If it’s those, then in principle the server would get the proxy’s IP-address, but Signal knows its users’ phone numbers in principle. They do anonymize it again if someone comes asking with a warrant but it’s again not something the client can mask for the user, unless the user takes separate steps of non-KYC SIM and Tor.
The threat model with Signal’s sever-side hardening gets murky if you want to build it around what they promise and try to deploy there. It’s not entirely clear to me which parts of the server’s functionality can the remote attestation check. So IMO it’s just easier to assume Signal is metadata private by policy, i.e., they could collect metadata if they wanted, but as per court docs, they don’t, and content private by design (client has excellent E2EE).
What do you think of the Android apps that “freeze” other apps from running in the background? Some OEM’s provide this where you could select an app that you don’t want running, enable your VPN/tor first and then unfreeze it?
SimpleX isn’t going to be used by me or my friends any time soon, no matter how much I care about my privacy. That’s the unfortunate reality we live in 
It’s just so funny to imagine that someone won’t use X tool/service because of Y feature (or lack thereof) despite Y being the least of your worries.
