You know, I've been having this conversation ever since PGP first came into existence. And much as I love the idea of encryption, and despite having invested lots of time in arguing for the right to encrypt and to share encryption algorithms etc. etc. I've always had to admit that if you're not a geek who loves computing for its own sake then encrypting all your email is a massive pain in the ass, whose costs substantially exceed the benefits for most people. Given that this argument has been going on for 20 years and that hardly anyone encrypts their emails on a daily basis, I'd say that the empirical evidence is that your solution, while clever, is Not Good because it doesn't meet people's actual needs.
Stop telling me why you like it and build something that's easy for other people to use. In your pursuit of technical excellence you are completely ignoring the importance of network effects on adoption and the disutility of standing out from the crowd by your use of super-solid encryption. Nobody wants to maintain a collection of public keys for every single person they know. No matter how bad things get politically there is not going to be a sudden mass awakening that will cause everyone to start using public key encryption for email, or we'd have already seen it take off like wildfire in politically repressive jurisdictions.
These are good points but I'll nitpick on one of them: we already maintain a collection of emails, phone numbers, etc. for everyone we know, and a public key is just one more data point in the contact list.
Whoah. Hang on there. Crypto specialists aren't the people lobbying loudly for PGP-encrypted email in 2017. They're the ones who made the world's most popular messaging application double-ratchet deniably encrypted by default without the userbase even noticing.
I'm sorry, I undermined my own point in pursuit of a witty syllogism and I don't want to give the wrong impression that the crypto in apps like Signal (which I prefer) is somehow second-rate.
WhatsApp isn't even open source, so no, it doesn't have reproducible builds. (Signal is open source and does have reproducible builds).
> Do we know that the app isn't storing the keys somewhere else?
No, but WhatsApp is still strictly better than either email or SMS. Even if you don't trust Facebook, that still only leaves three parties that can compromise the integrity of messages (you, the recipient, and Facebook), as opposed to the uncountable number of parties that can passively observe email or SMS in transit.
I'm hoping Apple will lead the way. They have enough devices in the wild that could force others to adopt over time. They did it with floppy drives and optical drives. I don't see why they couldn't do it with encrypted e-mails. With their stance on no backdoors, they could raise their privacy profile for consumers and increase sales of iPhones at the same time.
That the NSA cracked, backdoored, or intercepted most providers people trusted but couldnt beat GPG isnt an argument for GPG being secure? I think it's quite an endorsement for GPG given most people's adversaries will be weaker than NSA.
1) I don't take the snowden leaks as gospel, sorry.
2) Even if i did, "most peoples adversary" is a meaningless phrase at this point, and also used quite often as a rhetorical feint to take down someones arg. And given the profound unification of the security state across seemingly all lines, its also dead wrong. Technically everyones adversary is the NSA, as long as data is shared surreptiously and , more and more, openly and legally between TLA's, state, and local LEAs.
3) GPG may be an excellent tool, the first time you use it, but if you transmit anything encrypted you are automatically targeted, another point directly from the snowden docs, no? And since virtually no one is going to use one time devices and farraday cages unless your model of communication is "I just have to get this one message out, then I'm good" its worse than useless, given that it will only make you more of a target.
"1) I don't take the snowden leaks as gospel, sorry. "
Don't take them as gospel. That's faith. Review the evidence they're true from U.S. governments' reaction to them to what similar malware was found by third parties. Once evidence is in, then you have reason to believe them and then in stuff such as GPG by extension. And the leaks didn't say anything about faraday cages. Just that they had to rely on the extremely-limited resources of TAO... such as targeted attacks on specific sites/configurations/endpoints... if the target used something strong.
the idea that "evidence" comes from the govts reaction is just weak, its just a terrible argument. Extrapolating from that is also largely a mistake. Any number of possible interpretations of the docs themselves and the responses by the state could and have been made, each of which could point to mutually exclusive conclusions for toosl mentioned in the docs. This isn't a particularly novel argument (my arg, I mean) either. I wasn't pulling the faraday cages from the docs themselves, as well, I was positing that as an extreme example of data security.
Even the government stopped denying their truth, especially after the drip feed of info caught them in baldface lies on several occasions. Also, if the info wasn't genuine, why would they be so upset and calling Snowden a "traitor" etc?
Additionally, this wasn't even the first time any of us had heard about this sort of thing. Ever heard of Room 641A? The Snowden leaks aren't exactly implausible or unprecedented.
I don't know if you are genuinely unsure of their veracity or just making a rhetorical point, but it's not a point of contention in any serious debate forum I know of, even among intelligence community sympathizers.
Stop telling me why you like it and build something that's easy for other people to use. In your pursuit of technical excellence you are completely ignoring the importance of network effects on adoption and the disutility of standing out from the crowd by your use of super-solid encryption. Nobody wants to maintain a collection of public keys for every single person they know. No matter how bad things get politically there is not going to be a sudden mass awakening that will cause everyone to start using public key encryption for email, or we'd have already seen it take off like wildfire in politically repressive jurisdictions.
It's. Not. Going. To. Happen.