[Strint-attendees] Do we need a breakthrough in key management first?

[Johan Pouwelse <[e-mail]>]

On 17 February 2014 11:31, Carlo v. Loesch
<[e-mail]> wrote:
> And there are some papers, that propose fundamental
> rethinking of the way we run the Internet. Some are
> just rough ideas (26, 38) while others appear to have
[snip]
> 65: The Internet is Broken: Building a GNU Network
>
> Alas, I am biased. This is the paper I contributed to.
> For several years we have been working on design and
> implementation of an alternative Internet. Although it
> currently runs as an overlay network, it has developed
> protocols to replace DNS, X.509, BGP and various other
> insecure technologies. DHT-based cryptographic routing
> has matured in over a decade and grown well out of its
> infancy (just watch Tor). Why stick to horse carriages
> if there is a car waiting outside?

Key management seems to me the key problem that needs
a breakthrough and re-thinking.

How far can we travel without crossing that bridge?

But what does a solution look like. DHT-based overlays?
As an academic I think that the most promising direction
is a fully distributed key directory and reputation system.
Taking the idea of the "the people's CA" even further.
Everybody issues certificates in public and private.
End-users themselves store all data (so no DHT needed)
and successful user-to-user interactions feed into it.

We have Internet-deployed code within this direction.
It combines parts of the Tor specs with reputations
and removal of all central components/servers.
The aim is called the "Shadow Internet", see fresh I-D:

https://datatracker.ietf.org/doc/draft-pouwelse-perpass-shadow-internet/

For the past decade my university research team has worked
on this. We obtained 1.4 million installs, mainly due to the
Bittorrent-backwards compability. We now have implemented
the Tor tunnel specification.

Technical documentation: https://github.com/Tribler/tribler/wiki

Looking forward to discussion "alternative Internet" matters
with you at STRINT. (my whitepaper was too late, sorry,
became the above 19-page IETF draft)

Greetings, johan.