Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

You have a moral obligation to use crypto

blog.easydns.org

Post reply on HN
Print view View original

71–80 of 92 posts

Re: You have a moral obligation to use crypto

#71

Post by Joeboy »

peterwwillis wrote: post #51

If encrypting our communications were a moral imperative, wouldn't someone have suggested our telephone calls or postal mail be encrypted some time in the past 40 years? We've had the capability to do so as private citizens for years, but who the hell cares? And it's not like your bank is going to start sending your bank statements in the mail with a one-time pad. If we didn't have to think about crypto, everyone wou…

> If encrypting our communications were a moral imperative,

> wouldn't someone have suggested our telephone calls or

> postal mail be encrypted some time in the past 40 years?

Mass surveillance of those things wasn't really a threat, in most places anyway.

Re: You have a moral obligation to use crypto

#72

Post by dispense »

anovikov wrote: post #54
robertfw wrote: post #24

Reposting my comment on the blog: I can't comment on which encryption schemes are still strong, but one advantage to using protection for general use is providing strength in numbers. Right now, using Tor or strong encryption is a bright beacon saying "this person has something to hide!" and the NSA et al hone in by default. If, however, using Tor and strong encryption was the norm, it is easier for those who need it…

That won't work because Tor is too slow. I'd use only if i was a terrorist - it's too slow even for pedophiles. If we want people to use crypto en masse, we have to give them something faster. But the general idea is correct.

Perhaps it's been a while since you've last used Tor, but at least in my experience, the speed really isn't that bad, especially for tasks that aren't time critical. Speeds of hundred or several hundred KB/s are not uncommon. The latency is a couple of seconds at most, in most cases. You won't be streaming in HD over Tor, sure, but fetching your email over Tor is not significantly more tiresome than over a direct connection. There is usually no noticeable lag for instant messaging either. YouTube video streams in the Tor browser also works with no delays most of the time.

Re: You have a moral obligation to use crypto

#73

Post by mnw21cam »

GauntletWizard wrote: post #64
boronine wrote: post #63

Earlier quoted context omitted.

Does SSL not use a key exchange algorithm that ensures that a MITM proxy be useless?

Even without the CA system, SSL (and TLS) is secure against _passive_ MITM attacks; That is, if the attacker cannot alter the data-stream. This is still reasonable for many simple links, but for internet, it is not. If you've got no verification of the opposite party, literally any computer could be on the other end. The key exchange mechanism is secure, certainly, but any computer can execute it. The attacker can ev…

A "passive MITM" attack isn't MITM any more - it's just interception. MITM means that the man in the middle is able to control the data stream that both ends see.

Re: You have a moral obligation to use crypto

#74

Post by Zigurd »

sgift wrote: post #67
Zigurd wrote: post #58

Earlier quoted context omitted.

Of course Google deserves praise for that and 97 other things. But it is completely unapropos to what the article posted here is about. It does not really have an effect on dragnet surveillance because Google can be ordered to simply allow the government to let it happen. There is some value in that foreign governments with their own NSA wannabes could get thwarted.

The problem is that Google would probably closed down if they implemented security which cannot be circumvented, even with a warrant, because they would break the law. There are many laws and long standing rulings which force companies/people/.. to cooperate with law agencies. So, as it has been stated many times before, the solution for these things is political, not technical. Law always trumps programming and law…

What law would that break?

Re: You have a moral obligation to use crypto

#75

Post by hyaline9 »

sparkie wrote: post #13
Karunamon wrote: post #3

Earlier quoted context omitted.

As usual, misconfiguration causing scary warnings, useless to the end user, but the connection is still encrypted. I really wish we'd divorce the identity assurance part of PKI from the encryption part. I have no idea how it would be done, but.

Moxie introduced http://www.convergence.io/ a while ago now ( https://www.youtube.com/watch?v=8N4sb-SEpcg ), which could offer significant advantages over the PKI if it became widely adopted. Although it's not an end-all solution to identity, it's a step in the right direction (using web-of-trust ideas).

Unfortunately this project seems to have stalled; according to https://github.com/moxie0/Convergence the code hasn't seen significant updates since 2011.

A similar more up-to-date alternative is Perspectives: http://perspectives-project.org/

Re: You have a moral obligation to use crypto

#76

Post by oafitupa »

tptacek wrote: post #23
Zigurd wrote: post #20

The Internet services who are complaining should start coding, instead, or in addition to complaining. Secure key exchange, secure real time communication, secure storage, and secure email payload would blind the surveillance state. All surveillance states. Even if you think your own surveillance state is less than harmful, there are dozens of others which are patently evil or thoroughly corrupt. And yet, while we do…

Google has done more to securely encrypt Internet traffic than any other company in the world. Among other things, they are the pioneering standard-bearer for ECC forward secrecy and for certificate pinning, the two most important Internet encryption advances in the last 10 years.

Says the Google employee...

Re: You have a moral obligation to use crypto

#77

Post by x1024 »

anovikov wrote: post #54
robertfw wrote: post #24

Reposting my comment on the blog: I can't comment on which encryption schemes are still strong, but one advantage to using protection for general use is providing strength in numbers. Right now, using Tor or strong encryption is a bright beacon saying "this person has something to hide!" and the NSA et al hone in by default. If, however, using Tor and strong encryption was the norm, it is easier for those who need it…

That won't work because Tor is too slow. I'd use only if i was a terrorist - it's too slow even for pedophiles. If we want people to use crypto en masse, we have to give them something faster. But the general idea is correct.

Wait, did you just call yourself a pedophile?

Re: You have a moral obligation to use crypto

#78

Post by sparkie »

sparkie wrote: post #13
Karunamon wrote: post #3

Earlier quoted context omitted.

As usual, misconfiguration causing scary warnings, useless to the end user, but the connection is still encrypted. I really wish we'd divorce the identity assurance part of PKI from the encryption part. I have no idea how it would be done, but.

Moxie introduced http://www.convergence.io/ a while ago now ( https://www.youtube.com/watch?v=8N4sb-SEpcg ), which could offer significant advantages over the PKI if it became widely adopted. Although it's not an end-all solution to identity, it's a step in the right direction (using web-of-trust ideas).

@hyaline9, you're shadow-banned for some reason, people without showdead on will not see your replies, and we can't reply to you. I'd probably make a new account given that it was only created recently.

On perepectives: Convergence is actually based on perspectives with some improvements regarding privacy and latency (which are discussed in the youtube video). Are you aware if any of those improvements have been adopted into perspectives or not?

Re: You have a moral obligation to use crypto

#79

Post by dTal »

mnutt wrote: post #26
GauntletWizard wrote: post #4

Earlier quoted context omitted.

HTTPS Encryption is virtually useless without the identify verification part. Anyone can run a valid HTTPS server with a self-generated public key. Anyone could then place a MITM, and without the identity bit, you're just as compromised. If we had dropped the identity bit, every ISP would be running a MITM proxy, because they want control. Already, plenty of businesses enable poor hygiene by including transparent squ…

I was familiar with the reasons why using self-signed SSL didn't buy you anything, but hadn't really put together how it would completely weaken the whole ecosystem until now. "Oh, that's just Verizon MITMing me like always..."

Self-signed certificates are perfectly useful. Weaker security is still better than no security at all. For example, even unauthenticated, SSL protects you from log analysis after the fact. MITM is not the only threat model.

If it's a personal page (say, a public-facing login to your NAS) you can always check the SSL fingerprint too. You don't have to remember the whole thing, just remember a few characters and check that they're right.

Re: You have a moral obligation to use crypto

#80

Post by andrewaylett »

jedbrown wrote: post #47
jimmaswell wrote: post #34

Earlier quoted context omitted.

Do email sites not already use HTTPS?

Transfer between servers is over SMTP, which is not encrypted by default and almost always subject to degradation attacks (the MITM can claim not to support STARTTLS). In any case, the user cannot verify or demand that STARTTLS be used and many hosts don't support it at all. Meanwhile, PGP does not require any trust of intermediate entities, but relatively few people you might want to email have PGP keys and even tho…

Note that this is indeed almost always, rather than actually always. My mail server is configured to refuse to send to Google (and a few other hand-picked sites) unless protected by SSL. The converse is, unfortunately, not yet true.

DANE[0] should allow me to put DNSSEC secured records in place to tell Google (et al) that they should always use SSL to talk to me too, but I don't have DNSSEC support for my domains yet and my version of Postfix isn't new enough to be able to use it for outgoing mail either.

More frustratingly, many popular email providers (for example, BT) don't support SSL at all. That lowers the bar for eavesdropping from active MITM to passive listening.

0: http://www.internetsociety.org/deploy360/resources/dane/

Post reply on HN