[liberationtech] Why are we leaving mesh networking to FireChat?

[carlo von lynX]

On Sat, Jan 17, 2015 at 06:24:42AM -0500, Rich Kulawiec wrote:
> I rather expect than in another generation or two the entire obsolete
> closed-source ecosystem will be viewed as an unfortunate aberration
> in the evolution of computing.  This will happen whether anyone wants
> it to or not, because it's going to be *necessary* for it to happen
> in order to ensure privacy, security, and integrity in computation.

Wonderful. I love your entire post, Rich.
Inspires me to apply the thought to a pressing issue.

Over a decade of research in mesh networking has happened.
In projects like Netsukuku, Freifunk but also the frequently
(sorry) cited GNUnet. Why the hell is the open source community
still to the 99% focused on server-connectivity-dependent
technologies, leaving the next generation of activist tools
ONCE AGAIN in the hands of proprietary offerings?

> And again, *in the context we are in here*, it's absurd to even suggest
> that closed source software should be on the table for consideration.

Here we are in libtech, for several years now, and we haven't
been able to provide a mesh networking chat tool to the folks
in Hong Kong and elsewhere?

Briar? ServalMesh? cjdns? secushare? net2o? COR? Zyre?
All prototypes? Unfinished pieces of a puzzle?

Yes, cjdns is actually a running piece of software.. but
it was done quickly, with no sybil attack protection in
place. Predictable paths, no metadata obfuscation at all.

On the other hand GNUnet, with so much resarch gone into 
every detail - but noone there to fix it up and turn it 
into an actual product. Where are all the hackers? Why are
they playing around with SMTP over Tor and similar useless
stuff? [1]

How likely is it that FireChat suffers from sybil attacks?
Do we have to wait until somebody shows us? Causing harm to
activists and population in the poorest parts of Earth? Can't
we give them something solid?

Watch those videos explaining sybil-resistant DHT routing.
Bart has really nice slides to illustrate how that works. [2]
Understand that THIS is how we should route the entire
Internet. THIS is the foundation of a GNU Internet, where
anything you do is end-to-end encrypted, authenticated and 
resistant to censorship.

Remember John Gilmore saying "The Net interprets censorship 
as damage and routes around it" ? Well, with these technologies
for the first time that is actually true.

So put your PHP and Ruby aside. We need P2P code development.
We need agnostic relay nodes, not servers. Tor was only the
beginning. We need to dig deeper. We need mesh nodes all over
the Internet backbone just as much as in the favelas and the
city center demonstrations.

This isn't fringe technology. This is the next generation. It's
the clean slate which has surveillance-resistance built-in.
I know something about scalability, and there is no reason to
expect this stuff not to scale to worldwide use. At least the
well done projects.

And we need it to be encrypted, not unencrypted like FireChat.
The reckless way that is implemented, it's only a question of 
time until people will call it BackfireChat.


> Those who see Stallman as a "crazy wing nut" have not been paying
> attention -- or perhaps lack the analytical capabilities required to
> comprehend what they observe.

Word.

> 	The greatest shortcoming of the human race is man's inability
> 	to understand the exponential function.
> 		--- Albert A. Bartlett

Epic grand sigh.


[1] The "De-anonymizing Social Networks" attack described
    in the paper by that name is probably applicable to
    any popular XMPP or SMTP server, no matter how much
    it is kept behind Tor.
[2] http://youbroketheinternet.org/#30c3meshnet

-- 
	    http://youbroketheinternet.org
 ircs://psyced.org/youbroketheinternet