hello again

[Zooko <zooko@zooko.com>]

Folks:

I've been away from MixMinion for around two weeks now due to travel, others 
projects, and a bout of the flu.  The last thing that was happening when 
I left was the development of the "swap" and "strip" variants of the two-header 
approach to "tag-prevention with forward-backward indistinguishability".

I had been planning to write something about that, but perhaps the time has 
passed.

My point was going to be that my old "two payload" approach was actually 
deserving of consideration versus the two-header approach, because it isn't 
clear whether using twice the lower-level bandwidth for each message is 
actually more or less costly than using twice as many hops.  But my over-all 
leaning was more towards the simple old "one payload, one header, 
distinguishable forward-and-backward" design.

I haven't seen an overall statement from anyone else about the issue yet.  
Have y'all settled on one technique for the tagging issue?

I wonder if we're really going to get around to writing up a document about 
all these "paths not followed".  So far the tag-prevention issue has been the 
most difficult one for me personally to decide.

Regards,

Zooko

Zooko.Com -- Security and Distributed Systems Engineering