4 ms·

I wish there was something like Bitcoin for messaging. Users need a better way to prove their identities (or hide them, depending on the situation) and be able

I wish there was something like Bitcoin for messaging. Users need a better way to prove their identities (or hide them, depending on the situation) and be able to verify peers. The current hierarchical security system is probably destined to require money, but the free alternatives make me uncomfortable because they tend to need private keys. What I really want is an open version of Skype where the private key can always be generated by the user's mind, overcomes all the limitations like NAT that make users second class netizens, has no central tracker, and (optionally, but possibly required to work around the limitations) allows a certain amount of traffic to be forwarded like Tor. I don't see how we can have a conversation about who pays for messaging when the underlying infrastructure for truly secure online communication hasn't yet gone mainstream. I don't really trust Freenet or Tor and can't quite figure out how come, so maybe we could start there and figure out why that is. And a billion dollars for anything messaging-related completely blows my mind. Most of the world-changing software was developed for a few thousand dollars by a couple of people working in their spare time. To me, paying that much completely misses the point and probably hinders progress by locking capital up in niches that exclude developers who are working towards creating a non-scarcity based economy.

ignoramous 12y ago
https://tox.im https://tox.im ?
- zackmorris 12y ago
  That's really cool, I like how you just give the other user a Tox id token and it sets up the secure channel. I wonder if that's a general approach for secure communication, because as long as it's sent through another secure channel (that may use the traditional SSL certificates from an authority) it could be used to avoid a MITM attack. Maybe a way to do this in the future with an insecure channel is to give the peer your contact token using only context that both of you know. So the above token could get converted to a human-readable string like "that place we met that first time and what you said to me with 12345abcdef on the end" and the other party would have to know you to figure out how to find you. Proving who you are beyond that point would probably involve some back and forth questioning like in spy movies or a video call (which could be faked in 5-10 years with computer rendering) so shared history will be key. It already sounds like it can run over Tor, but it would be nice if it had a standard socket wrapper that could be run over UDP NAT tunneling to look like an SSH VPN. Maybe WebRTC could be integrated with it somehow. I just mean that I'd like the negotiation phase to be separate so that if it works, you're just presented with a secure socket.
maxerickson 12y ago
In 5 years time a widespread conclusion might be that Telegram is the Bitcoin of messaging. (But probably not in a way where anybody is glad about it)
Nutomic 12y ago
I have developed an instant messenger for Android called EnsiChat [1]. For identity, I use an RSA key pair. For key exchange, devices send their public key to direct neighbors, and you can visually check that the key is correct. I have also thought about implementing transport over internet (right now it's only Bluetooth). But that would require a lot of effort. [1] https://github.com/Nutomic/ensichat https://github.com/Nutomic/ensichat
pain 12y ago
http://Matrix.org http://Matrix.org makes e2e pgp encryption of audio/video/text chat friendly and federated messaging. Bitcoin is on their radar for identity proof, as one of many ways to verify https://matrix.org/docs/spec/#n-factor-authentication https://matrix.org/docs/spec/#n-factor-authentication and could use your help

ignoramous 12y ago

https://tox.im https://tox.im ?

zackmorris 12y ago
That's really cool, I like how you just give the other user a Tox id token and it sets up the secure channel. I wonder if that's a general approach for secure communication, because as long as it's sent through another secure channel (that may use the traditional SSL certificates from an authority) it could be used to avoid a MITM attack. Maybe a way to do this in the future with an insecure channel is to give the peer your contact token using only context that both of you know. So the above token could get converted to a human-readable string like "that place we met that first time and what you said to me with 12345abcdef on the end" and the other party would have to know you to figure out how to find you. Proving who you are beyond that point would probably involve some back and forth questioning like in spy movies or a video call (which could be faked in 5-10 years with computer rendering) so shared history will be key. It already sounds like it can run over Tor, but it would be nice if it had a standard socket wrapper that could be run over UDP NAT tunneling to look like an SSH VPN. Maybe WebRTC could be integrated with it somehow. I just mean that I'd like the negotiation phase to be separate so that if it works, you're just presented with a secure socket.

maxerickson 12y ago

In 5 years time a widespread conclusion might be that Telegram is the Bitcoin of messaging. (But probably not in a way where anybody is glad about it)

Nutomic 12y ago

I have developed an instant messenger for Android called EnsiChat [1]. For identity, I use an RSA key pair. For key exchange, devices send their public key to direct neighbors, and you can visually check that the key is correct. I have also thought about implementing transport over internet (right now it's only Bluetooth). But that would require a lot of effort. [1] https://github.com/Nutomic/ensichat https://github.com/Nutomic/ensichat

pain 12y ago

http://Matrix.org http://Matrix.org makes e2e pgp encryption of audio/video/text chat friendly and federated messaging. Bitcoin is on their radar for identity proof, as one of many ways to verify https://matrix.org/docs/spec/#n-factor-authentication https://matrix.org/docs/spec/#n-factor-authentication and could use your help