Re: More discussion of header swap efficiency

[Roger Dingledine <arma@mit.edu>]

On Fri, May 03, 2002 at 06:39:43AM -0400, George Danezis wrote:
> I buy your argument. We should delete that phrase, and insert your 
> explanation on why we do not need double length.
> 
> Yours,
> 
> George

Ok. So David -- at this point, I'm thinking having a second option
(distinguishable forward and reply) in the paper is just going to confuse
and distract from the the rest of the issues in the paper. Is there any
respect in which the distinguishable approach is better?

I'll write up (the rest of) the header swap approach Friday night or
Saturday. I've been stalling to see how else you might break it. So at
this point I'm hoping either to hear about why distinguishable is better
than header swap, or why header swap is broken.

Thanks,
--Roger