Re: [tor-talk] phantom protocol

[grarpamp]

On Sun, Dec 14, 2014 at 4:22 PM, Jonathan Wilkes <janc...@yahoo.com> wrote:
> Does anyone here know about the Phantom Protocol:
> https://code.google.com/p/phantom/
>
> It looks like it's abandonware, although someone sent a message to the list
> that they had done some tests with virtual machines.

The HESSLA is problem of untested utopian clickwrap that
criminals and govts will ignore, and that users and authors
don't know or want to deal with, so it needs relicensed to the
familiar BSD (two or three clause, like Tor uses) before it is
likely to attract new developers.

Phantom does run, but may need some fixes to compile,
and the platform specific tun(4) code needs more work.
One contributor put in some good updates, but against an
old rev, so that needs merged.

> My main question is this: how does it bootstrap?

Seeded with nodes you know and configure, learned thereafter.

> I'm asking here because it does onion routing, so maybe someone here has
> investigated its design and/or implementation.

Fully DHT decentralized routing, no exits, presents its
anonymous network over a local IPv6 interface thus works
with all your favorite IPv6 enabled software. Probably most
similar to CJDNS as far as using it goes.

You'll want to read their whitepapers and watch the video.
Both of which should be linked above.
It's worth looking at.

[grarpamp]

On Mon, Dec 15, 2014 at 7:42 AM, str4d <st...@i2pmail.org> wrote:
> As far as the design goes, it is most similar to I2P's protocol. I
> have only read (most of) the original whitepaper[0] so far, and AFAICT
> at a high level it is almost identical to I2P, except for a few
> specific points (e.g. using IPv4 for identifying anonymizing paths,
> and bidirectional circuit-switched tunnels).

Yes the IPv6 interface is a specific point and a major feature. Like
CJDNS and OnionCat [1]... that gives the user access to whatever
apps they want to use over it, right out of the box. TCP, UDP, ICMP,
all of it, no problem. No porting, no waiting for someone to write an
app for the network from scratch, no need for socksv5.

I like forcing new dedicated apps (ideas) to appear [2], but not at the
expense of a variety of use / network adoption. Unless of course
the network is specifically designed for one thing. And with all
networks, using them securely is still up to the user as always.

[1] Tor-HS-v2 will break onioncat one-to-one addressing and thus be
no more useful than garlicat at that point. I think we need to come
up with *cat-v2 for those to regain one-to-one somehow.

[2] For example: I2P-Bote, Pond.

> It is an interesting protocol. I am reviewing the whitepapers, and
> intend to publish a comparison page on the I2P website[1] once I have
> a good understanding of the differences.

I'll add CJDNS somewhere when I can.