upvote
The message can't be intercepted in transit, since we are talking about spyware, I assume they get it from the device, hard to defend against that if they have access to your process' memory space.
reply
Certainly very hard to defend against that when the messenger you're using won't let you use a device you control.
reply
Surprising that end-to-end encryption doesn't really matter when you get into one of the ends.
reply
Even if you had to input your private key every time you wanted to read or send a message, having malware in your phone voids practically any form of encryption, because it has to be decrypted eventually to be used.
reply
not at all. there is no encryption that can save you when one of the legitimate participants is somehow compromised. doesn't even need to be a sophisticated device compromise, literal shoulder surfing does that too.
reply
[flagged]
reply
The parent said "it's surprising". It's not surprising.
reply
You're correct in the literal sense that they did say those words, but the entire comment clearly demonstrated a lack of surprise that reveals the opening words to be intended ironically.
reply
>The message can't be intercepted in transit

Lol, so like ... all encryption schemes since the 70s?

reply
They do have stronger schemes, which are called hash functions.
reply
What?

Hashing is not encrypting.

You can learn more about the topic here, https://www.okta.com/identity-101/hashing-vs-encryption/

reply
It's a joke, because hashing loses information, and thus the original is not retrievable, woosh
reply
Hashing is a part of encryption, maybe you are the one who needs to shore up on the topic?
reply
A good hash function is surjective. Encryption is bijective. They're very different things.
reply
Nice try. However, hashing and encryption are two different operations.

Load this page, https://en.wikipedia.org/wiki/Advanced_Encryption_Standard

Ctrl-F "hash". No mention of it.

Before being pedantic at least check out the url in that comment to get the basics going.

reply
This entire thread should be annihilated, but since you mentioned being pedantic...

You're correct that a pure encryption algorithm doesn't use hashing. But real-world encryption systems will include an HMAC to detect whether messages were altered in transit. HMACs do use hash functions.

reply
> What?

> Hashing is not encrypting.

> You can learn more about the topic here, https://www.okta.com/identity-101/hashing-vs-encryption/

Thank you for that link. Your original comment implied that Signal's threat model should have included an attacker-controlled end. The only way to do that is to make decryption impossible by anyone, including the intended recipient. A labyrinthine way to do that would be to substitute the symmetric-encryption algorithm with a hash algorithm, which of course destroys the plaintext, but does accomplish the goal of obfuscating it in transit, at rest, and forever.

reply
How is this related?
reply
I see there's some room for ambiguity.

See, https://en.wikipedia.org/wiki/Moxie_Marlinspike

reply
Apologies for being dense. Could you spell out how you went from Paragon Solutions to the Signal Protocol?
reply
I guess they've seen a Signal icon in the photo. Of course the interception is done locally on the phone (so it's basically "man-in-the-client" rather than a "man-in-the-middle"), therefore the Signal protocol is not really worth being mentioned as it has nothing to do with local interception.
reply
Cool, can you now show how the protocol has been broken? Lot of smart people would love to see your novel research.
reply
Yea I knew which Moxie it was but that didn’t help at all haha
reply
It’s performative security, when an app still requires a phone number, can’t have your own server, and all these audits are meaningless as you might have memory injected spyware later, it is NOT secure, never was.
reply