Re: Thoughts on the remailer network

[Nick Mathewson <nickm@alum.mit.edu>]

On Tue, 2002-04-23 at 02:44, Len Sassaman wrote:

Hi, Len!  Though it's been our long-term goal to get accepted as the
protocol for for MM3, I never thought we'd get so far this early! :) 

 [...]
> There were other issues we wished to address too, but those were the most
> pressing. For anyone truly interested, you can dig through the archives of
> the mixmaster-devel list on SourceForge (though there is a lot of "I found
> a memory leak" type noise on that list as well.)

Thanks; I'll be sure to check this out soon.

 [...]
> So, here's one way things could play out that would give you maximum
> adoption by the remailer community:

BTW, I concur that a RFC is a good place for our spec to wind up.

To summarize, you're suggesting that we go:

  1. Spec
  2. Mixmaster 3 supports spec
  3. Nymservers support MM3
  4. Nymserver users switch to MM3
  5. Type I remailers shut down; switch to Mixmaster 3.5.
  6. Type II clients get Type III support.
  7. Type II remailers shut down; switch to Mixmaster 4.0.

I'm going to suggest that the Mixminion Python implementation go out
somewhere between 1 and 3 as well.  That way, we'll know from the get-go
that we have compatible implementations.  (Naturally, if one of the
implementation projects gets stalled, the other can continue to move
ahead.)

I'd also like to suggest that we specify things in such a way that you
can get the same effect as Mixmaster 3 by running Mixmaster 2/3/3.5 (for
types I&II) along side Mixminion (for type III).  That way, people can
test out with both implementations of the type III protocol if they
would like.

 [...]
> All the while, MixMinion (which I understand you are doing in Python)
> should be in use, and hopefully be attracting new remops who will be
> running Type III only remailers. 

Indeed, one of the main goals for the Mixminion implementation is to get
as many nodes as possible by reducing the administrative complexity and
the legal burden (e.g. by making it easier to have a relay-only node).

The next step (which is still everybody's favorite research project) is
to make the system more tolerant of less-reliable remailers... but I
think that's not for v3.

>                                   I'd also like to see MixMinion-based
> clients for a wide variety of OSes, particularly ones like MacOS which do
> not have decent Type II client support, so that demand for Type III
> remailers is high.

I agree here as well; this is another one of our implementation goals. 
Leveraging existing Mixmaster-based projects could be invaluable for
client support too.

{Fortunately, Python on MacOS is rumored to work quite nicely.}

Yours,
-- 
Nick