Re: [dispatch] [Secdispatch] A protocol for anonymity

[Eric Burger <[e-mail]>]

I’m curious about the ‘main’ document. Is the document you mention below supposed to be at
https://messagevortex.net/devel/messageVortex.pdf?
If so, the document is broken. As in unretrievable, not a value judgement of the technical content. It may be that, too, but I’ll reserve judgement.

IMHO the world of anonymous communication is full of ToR knockoffs and protocol steganographies. I would be curious to learn how Message Vortex is different and advances the state of the art. See, for example:
https://www.darpa.mil/program/resilient-anonymous-communication-for-everyone

A *very* quick read of the I-D indicates there is a lot left unspecified. For example:
- key distribution is done by magic
- accounting is disturbing: that implies nodes have a *lot* of tracking capability
- I’m not sure blending will work, but that will require a bit more reading on my part (anyone else want to chime in)?
- the hand wave of Dead Parrot is fatal: one can tell these are Message Vortex messages by simple inspection. That defeats the purpose, no?

Finally, ASN.1? Really? I suppose we could use X.500 for message transport :-)


> On Mar 11, 2019, at 1:31 AM, Martin <[e-mail]> wrote:
> 
> Hi Rich
> 
> Thanks for your concern. I understand where it is coming from. The
> reason why Security Considerations are so small is that there is another
> (academia) document dealing mainly with the security of the protocol.
> Although the thesis has not been published (still needs lecturing, a
> couple of attentive eyes questioning everything, and extension of some
> sections), it is linked in the document, and you may read it. I felt
> that including this into the RFC is too much by far. I, therefore,
> referenced it in the first paragraph. Do you think that this was the
> wrong decision?
> 
> You made definitely a valid point: Why believe a single person (in terms
> of an administrator of a system or writer of a paper)? My reply is: You
> should not. Read it and build your own opinion. There are a lot of minor
> flaws in the protocol depending on how your adversary looks like. A lot
> of "do's" and "donts's." If you are looking for the "why's" then reading
> [MVAnalysis] is mandatory.
> 
> Another main problem is censorship resistance. The more you regulate the
> behavior of the node, the easier it is to identify nodes (at least in
> some points). I, therefore, tried to integrate all absolute mandatory
> rules and to leave freedom to all other rules. I am sure willing to
> change that decision if you think it should be entirely included in the
> document instead of linking, but the document is already far from being
> lightweight.
> 
> Regards
> Martin
> 
> 
> Am 11.03.2019 um 03:20 schrieb Salz, Rich:
>>>   https://datatracker.ietf.org/doc/draft-gwerder-messagevortexmain/
>> 
>> I am VERY concerned about publishing something that claims to be a protocol for anonymity that has no security analysis and seems to be the work of one person.
>> 
> 
> _______________________________________________
> dispatch mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/dispatch