[Cryptography] New White Paper: GhostLine - Information-Theoretically Secure Multi-Party Chat

[Pierre Abbat]

On Saturday, September 6, 2025 10:49:49 AM EDT Ron Garret wrote:
> I'm posting this as a public service to the crypto list.  I looked at this
> paper so you don't have to.  The paper appears superficially plausible, but
> the general heuristic that any mention of a one-time-pad is a strong
> indicator of crackpottery applies here.  AFAICT on a cursory reading, the
> paper isn't wrong, it just leaves out one teensy little detail: key
> exchange.  There is a reason that OTP is never used in practice despite
> being information-theoretically secure.
> 
> There are additional clues: the "From" header lists the senders name as
> "Ferecides de Siros" [1] but the message body and the paper both say that
> the author's name is "Hitokiri Battosai" [2], and the "about me" page on
> the web site listed in the contact information in the paper says "I'm
> Aristoh4ck8r, the enigmatic force behind the curtain."

All of these were my concerns as well when I gave it a quick glance, except 
that I didn't get as far as "Aristoh4ck8r". (I'm moving and was too tired to 
think of crypto; also, Ferecides posted it at a time that here is the night of 
the Sabbath, and I had to get enough sleep to not nod off too much in church.) 
I just cloned the repo and looked at the code, of which there is a lot less 
than I expected, and have some more questions: How do you get truly random 
numbers for the OTP? How do you distribute the OTP? What happens if Alice 
sends Bob a message at the same time that Carol sends Dave a message? Can they 
use the same piece of OTP, thus producing depth?

Pierre
-- 
Don't buy a French car in Holland. It may be a citroen.