Re:MinimaLT: Minimal-latency Networking Through Better Secur...
by ableal on Thursday May 23, 2013 @03:06AM
(#5755672)
Worth a look (authors include D.J. Bernstein):
We describe here MinimaLT, a secure network protocol which delivers protected data on the rst packet of a typical client-server connection.
MinimaLT provides substantial protections and is extraordinarily simple to configure
and use. In particular, it provides cryptographic authentication of servers and users; encryption of communication;
simplicity of protocol, implementation, and conguration;
clean IP-address mobility; and DoS protections"
Re:MinimaLT: Minimal-latency Networking Through Better Secur...
by mrmekon on Thursday May 23, 2013 @06:01AM
(#5756009)
If I hadn't seen DJB's name on this, I would have read the abstract and thought "bullshit!" Now I just assume it's the most brilliant protocol ever written, and the source code is completely illegible :D
Re:MinimaLT: Minimal-latency Networking Through Better Secur...
by Historiopode on Thursday May 23, 2013 @06:30AM
(#5756096)
While I lack the skill set to fully appreciate the paper, I must say that it is very refreshing to see innovative proposals on this kind of all-important foundational elements.
These are the kind of submissions that make HN worth visiting, for all its problems. ;)
Re:MinimaLT: Minimal-latency Networking Through Better Secur...
by lvh on Thursday May 23, 2013 @04:17AM
(#5755802)
It is provided, just typeset weirdly. UIC is centered between the first four authors (Petullo, Zhang, Solworth, Bernstein). AFAIK all four of those are at UIC :)
According to the author's own website[1], his position at UIC is ongoing, as well.
We describe here MinimaLT, a secure network protocol which delivers protected data on the rst packet of a typical client-server connection. MinimaLT provides substantial protections and is extraordinarily simple to configure and use. In particular, it provides cryptographic authentication of servers and users; encryption of communication; simplicity of protocol, implementation, and conguration; clean IP-address mobility; and DoS protections"
These are the kind of submissions that make HN worth visiting, for all its problems. ;)
We plan to soon release Ethos [experimental robust-security OS] and our Linux MinimaLT implementation as open source software.
(Side note: I am not sure why it was downvoted, I apologize if my question is offensive/problematic for people. )
According to the author's own website[1], his position at UIC is ongoing, as well.
[1]: http://cr.yp.to/positions.html