NSLs, gag-orders, code-changes, coerced backdoors - any tech response? (Re: Lavabit and End-point Security)

Eugen Leitl eugen at leitl.org
Sun Aug 25 01:04:39 PDT 2013


On Thu, Aug 22, 2013 at 01:51:16AM -0400, grarpamp wrote:

> I like the decentral model. But I wonder about how to affirmatively
> deny an influx of attacking nodes overtaking the network. It surely
> cannot be relegated to the simple user? So that seems 'hard' to me.

You need each node reputation stored in a global distributed
tamper-proof publishing system, obtained and acted upon by global
quorum. This is not easy, but Bitcoin and Tahoe LAFS show how
to build a more trusted network from untrusted components.

> For example, I think Tor may remain centralish rather than pure
> dhtish for that purpose. But what if the centrality was undertaken

If you want to scale to millions if not billions of nodes, what
are your options?

> anonymously by some voting humans (or their analytic nodes). Their track
> recourd could certainly be public yet anonymous therein. You would
> at that point be trusting/subscribing their record, purely, as opposed to
> dht or some other means, purely. What would p2p-hackers@ have to
> say on this?



More information about the cypherpunks mailing list