Wikileaks is the Endgame

grarpamp grarpamp at gmail.com
Thu Jun 30 01:18:30 PDT 2016


On 6/30/16, Mirimir <mirimir at riseup.net> wrote:
> Yes, far more papers than implementations.

Papers are a legitmate validation route before investing
in implementation. Though when faced with many
rough equivalents, moving to code is the ballsy right move.

> Once you have something that
> works well enough, or that people are at least so convinced, new ideas
> don't attract much interest. But maybe that's changing now.

We know fossil fuel is unsustainable at current rates.

> Why did Tor go with circuit switching?

You'd have to ask Paul and the historical anonbib on that.

> Was it mostly about efficiency?

Perhaps. And or prevalance of TCP / parts of historical
thinking telecom models.

> Also, is it easier to bury packet switching in chaff?

Which is more programmable and pluggable?
Which is harder to emulate with the other?
Circuit switching, packet switching, packet switching with MPLS.
Why doesn't Ma Bell have 1:1 hardlines from NYC to Delhi?
And does it matter to you?

> Which ones, in particular?

Pick any of them and folks will say some parts
are really cool and effective, other parts not so much.

> But it's still obvious that you're up to something.

It seems very hard to hide that fact. But so long as free
speech and or via crypto is upheld, actual fact of usage
doesn't seem that bad. If not held, you've probably got
bigger problems to where Libtech has to sneak in and
Arab Spring your ass out.

> And
> once they tweak your traffic, they can look at what other nodes start
> behaving differently.

Not if those other nodes refuse to talk with or pass traffic for you
until your traffic falls back in line with expected network params.
Think of it this way... TCP is nice because it silently and
flexibly adapts to network conditions. When your adversary
*is* the network condition, you may not want that feature.
Tor checks for invalid packets via the OS stack, cell format and crypto.
But it and others have zero visibility into me shaping your
throughput into nice waveforms and watching whatever else
your NIC does appear straight out the other side somewhere.
That's dumb as fuck.

> So I come back to the need for covert channels.

These are definitely useful. SDR, guerilla community
meshnets, sat transponders, tropo, bulk sneakernets.

> We have a very different Internet now.

What do the spies say? Does it matter?



More information about the cypherpunks mailing list