Hacker News new | comments | ask | show | jobs | submit login

I've honestly never really quite recovered from watching the dreams of the cypherpunks die so hideously and completely. I guess it's because the ideas / movement / whatever bubbled up when I was at that age when you think this is going to be the groundswell, my generation is going to cause a fundamental change in the world.

If you're unfamiliar, there was a strong meme in the late eighties through early/mid nineties among a certain set that the perfect storm of public key encryption (still wonderfully unbounded in our minds) and the emerging global network would be a nexus point for personal power in privacy, anonymity and security and in many real ways break down the bonds of the states. It's worth noting that this was about the time that the soviet union fell, and many in the know had gotten a first taste of global presence by hearing about the people in the streets via usenet before it made the news.

It wasn't that I was particularly a hard core believer or activist, at least compared to many I knew. But for those who understood what an immense impact the internet was going to have it seemed to everyone I knew - NSA, hackers, professors, that it was just how it would be. You couldn't hope to spy on pretty much anyone anymore when you could use perfect encryption to scramble a telephone call or an email. Kind of like when you knew everyone was about to have a touch phone.

I was ideologically aligned and mixed in such circles, nerds were still outcasts so not really too big a world, but my life was busy with other things - but I watched from a distance, fascinated with all the ideas and things to come. I'm not sure I've every really been more sure something was going to happen, at least to a very significant degree.

The government was sure too - that was when they came up with CALEA and people got upset but mostly scoffed - there was a real sense that they were just in their death throws.

Things got pretty busy, Internet boom. Company got bought by an agency, every big name anybody needed to be on the Internet yesterday. Was a blast though a bit of a blur - ended up in SF as the whole thing worked itself into a nasty hangover. Can't remember worrying too much about when the cypherpunks were going to win but still knew it had to be coming, err well it's just about adoption.

It really sucks to wake up after a bender and realize that you helped kill the dream that you were just waiting for someone else to make happen.

Working infosec as california recovered put me face to face with reality pretty early in this cycle. Not only was the thing I was so sure of totally not how it went down, with shift from relatively petty financial fraud and wankers to states and srs.bsns abandoning defense to focus solely of offense it's been very hard to square. It's hard to believe many people ever feel so sure about something that turns out so absolutely opposite.

Fuck, at least nobody killed rms.




Well, don't completely give up hope just yet! All the tools needed to create the world you mentioned exist today. We have good open-source encryption that can be applied to all sorts of data and hide your tracks if you really want to. What we don't have(yet) is something so easy that the general public can do it. Like, tap a padlock icon on your cellphone and suddenly anything your phone does is encrypted and only the receiving party can decrypt it. GPG/PGP-plugins could be added to just about anything. Hang in there and keep up the fight. Keep explaining your point of view to anyone who will listen. There have been times in history where seemingly-invinsible corporations and/or groups of very privileged people get overturned when the general public "wake up" to what's going on. Even what's going on in North Korea won't last forever, there will be someday that one major incident that domino-effects into the downfall of that whole system.

"And never forget, the internet only knows what you tell it... more or less" --me!


>>> tap a padlock icon on your cellphone and suddenly anything your phone does is encrypted and only the receiving party can decrypt it.

Or so you think until it turns out an amendment to 2000-page farm appropriation bill actually mandated a government backdoor to be installed into any phone legally sold in the US, and 100% of US providers implemented in 5 years ago. And this backdoor is accessible without warrant since you communicate over public airwaves so you have no expectation of privacy.



do you have details on this, not being snarky, i genuinely want to know.



> Like, tap a padlock icon on your cellphone and suddenly anything your phone does is encrypted

Not good enough: you have to think of activating it. And even if you do, most traffic will still be unencrypted, making it easier for spies to tell who may have something to hide, and when they do.

To have real good, actual privacy, everything should be encrypted by default, the internet itself should be a giant scrambling overlay network such as Tor, and people should have symmetric bandwidth to encourage decentralization —no more need for YouTube.

I don't see it happen in the following decades.


I think that's what screwed it all up, in the end: lack of symmetric bandwidth. Without symmetric bandwidth, everyone needs to talk to centralized services and servers in order to pass any data around larger than an email attachment. Those services and servers become power centers, which attract surveillance and corruption.

A really safe internet has to look more like BitTorrent and less like YouTube.


Also, pervasive NAT and dynamic IPs. David Reed mentioned somewhere that he'd argued against 32-bit IP addresses and lost.

There've been powerful incentives for software on servers; I think the above got in the way of p2p getting much of a foothold to develop its own advantages.


There are initiatives going in this direction:

http://thefnf.org

http://projectmeshnet.org


>everything should be encrypted by default

Regarding phones, this is already the case with iOS. The Full-disk and Full-filesystem encryption mechanisms appear to be fairly/very strong. I believe since Android 4, full filesystem encryption has been supported, but I'm not sure if it's as well-integrated as on iOS.

Apple has made an effort (although an imperfect one) to make text messaging secure by default.

Obviously Apple screwed up pretty badly by making all this stuff closed-source, and it's probably full of vulnerabilities, but the reality is that this seems to be, in practice, enough to thwart LEO attempts to surveil users of iOS devices.

I think we're on the right track.


Encryption doesn't mean:

- that it's implementation secure (for example Android FDE is trivial to crack for us, imagine what a joke it is for the NSA)

- that it's algorithms are secure (gone are the days where the NSA would warn us DES is broken to help US companies)

- that the data is never stored or sent in clear - the system has full access to the data, in clear, when its running.

- that there is no backdoor. each step of the implementation can have relatively hard to spot backdoors. Specially in proprietary code.


> Android FDE is trivial to crack for us

What vulnerabilities exist with the full disk encryption on android that make it insecure ?


Likely stuff like http://appleinsider.com/articles/13/03/07/researchers-bypass... - basically, the encryption is fine, but you need somewhere to store your keys. "Somewhere" ends up being RAM, which is usually not very secure.


What's the problem with most traffic being unencrypted? If you have something in particular that you want hidden, it's possible to make sure that it gets hid. What more do you need?


You not to know I'm hiding anything at all.


Exactly. They may not know what you're hiding, but if they see small portions of your internet traffic are encrypted using a scheme very different from your regular traffic, flags are raised. Then it's simply a matter of sending an NSL to whoever they need to in order to get the content of the message.


>Then it's simply a matter of sending an NSL to whoever they need to in order to get the content of the message.

If its done right, there are only 2 people capable of getting the content of the message: you and the intended recipient.


If the norm was that everybody sent their snailmail on postcards, the ones in envelopes would be suspicious.


> What we don't have(yet) is something so easy that the general public can do it.

Companies are starting to realize that providing a good experience is the most important thing your company can do to stay relevant. Dreams often die in execution -- but I think we're getting to a point where a group of dedicated individuals focused on creating an exceptional Internet experience built on privacy and encryption really could make something happen.

I used to use GPG on Linux and OS X. I tried so very hard to keep using it. But it was the biggest pain in the ass, so eventually I just gave it up. I pulled up my key not long ago and had totally forgotten my passphrase.

The devil is in the details. Creating something familiar, something usable, and something that the average person would actually want to use is the part we need to get right.


Have you used the new GPGTools client for OS X yet? It's the easiest thing in the world to encrypt emails. I don't even have to put a second thought in. OS X also supports S/MIME out of the box, which is just as easy (if not easier) to use after you get it set up.


I used something similar back in 2007 and it just became a pain. It was never unusable -- lots of people do it -- it was just never truly seamless for what I saw as decreasing value for the minor frustration. The experience of using this kind of technology should be as embedded in everyday life as keeping your keys with you all the time, in my opinion. That's how you get people to do it more.


We still haven't figured out a way around rubber hose decryption, and the sad truth is that we're more likely to draw attention to ourselves by encrypting everything in personal email than we are by trying to fly under the radar.

I'm still terrified of a second McCarthyism.


There is no perfect way around rubber hose decryption, but we've some defenses. Off the top of my head, some significant ones:

    1) Proper anonymity, so they don't know who to beat.
    2) Deniable encryption.
    3) Steganography.
    4) (with 2) Sacrificial data of less significance to "give up" after sufficient beatings.
    5) Social norms against beatings and similar coercion, extending to extreme circumstances.
    6) Governmental transparency.
Neither individually nor collectively are these perfect security (and some are only relevant to certain circumstances) but they help to limit it.


7) destroy the data (or key) in case everything else fails


> Like, tap a padlock icon on your cellphone and suddenly anything your phone does is encrypted and only the receiving party can decrypt it.

That's not enough. They still know who you talk to, where you are and just about everything except the contents of your conversation.


> I've honestly never really quite recovered from watching the dreams of the cypherpunks die so hideously and completely.

This is really what the rage on the internet is about. People like to pretend they're mad at the government overstepping its Constitutional boundaries, but what they're really mad about is the failure of their attempt to re-litigate the division of power between government and the people.


I think that's true, though there's a third axis, companies, which is a big part of that failure. One major change in the tech scene over the past few decades is that a significant proportion of us effectively went over the other side: many techies now work for companies whose goal is, as with the government, to collect data on people, construct profiles, and share it around as convenient. That left rather fewer (though still vocal) people working for the opposite goal of anonymity, non-trackability, and flexible/modifiable pseudonymous identities. It also made it much easier for the government to piggyback on that tracking infrastructure we ourselves are building.


> It also made it much easier for the government to piggyback on that tracking infrastructure we ourselves are building.

Not just easier: we made it possible.

The government couldn't have created Facebook on their own. But now that it exists, it's an intelligence agency's dream come true.


That's it in a nutshell.

Every time we see another consumer web startup which relies on advertising or mining user data, we see yet another nail in the coffin in freedom and privacy for users. But, that's okay, because we're totally killing it, and that bridge round is coming any time now to keep us in our expensive lofts and designer foods.

Way to go, folks. Hope it was worth selling out the rest of your fucking race.


Yes because I really did hold my grandmother at gunpoint as she gave me the 14th update in the last hour about her stupid cat.


It's the conflict between two different paradigms for the internet. Cyperpunks want to think of the internet as a tool for trading ideas, potentially subversive ones. Most people think of the internet as a social or commercial space.


Nicely put. When viewed that way, the internal "nymwars" at Google looks like a pretty direct culture clash between those paradigms: should G+ allow pseudonymous profiles to protect free exchange of ideas (possibly including stigmatized or even dangerous ones), or should it insist on real names, to reduce spam, trolling, and other trouble-making, and improve the quality of the data collected?

It seems Google had a significant number of employees in each camp. But the market logic was pretty firmly in one of those camps and not the other...


That's a great point, and it points to a weakness that most privacy advocates don't really address very well: knowing more about you allows some companies to do things better for you. They have to answer the question, how do we provide the same awesome services while also preserving privacy?


> Fuck, at least nobody killed rms.

Oh, that's probably partly because that wouldn't help its detractors any one bit: as a dead martyr, RMS would be more powerful than as a living bitter old figurehead. (Disclaimer: I know nothing about RMS' actual mood.)


Even in those times, it was realized that there are 300,000,000 Americans. That there was no way that any techonology, including today's, could begin to monitor any but the crudest details about all the traffic so many generate. That such an insurmountable problem can be made even harder with encryption, phoney identities, phoney traffic, and dozens of counter-counter-measures. Let alone be privy to what we do and say 'out-of-range'.

Only the monitoring computers have the time and patience to look at as much as they can ... and they can't parse text sentences (let alone voice comms) well enough to do anything but scrutinize for a few common terms ... let alone nuances (seen Google translate?). A couple of back-of-the-envelope calculations will demonstrate that to anyone.

The agencies and the corporations know that but they refuse to cop to it, possibly because it's so obvious that all they can do is -pretend- to be able to monitor a significant fraction of it all. Maybe because pretending is the only hope they've got left.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: