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

Ferecides de Siros filosofarte at protonmail.com
Mon Sep 8 20:51:27 EDT 2025


On Monday, September 8th, 2025 at 16:14, Jon Callas <jon at callas.org> wrote:

> If you're going to use Mechanism X to transmit your pads, why not use that same mechanism to just transmit the message?
> 
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Hi Jon,

Thank you for this brilliant historical perspective - it's truly a masterclass in the practical 
realities of information-theoretic security. Your analogy of Sauron's seduction perfectly 
captures both the allure and the operational challenges of one-time pads.

You've articulated the fundamental paradox elegantly: if we have a secure channel for pad 
distribution, why not use it directly for messages? GhostLine indeed targets the specific 
niche where pad distribution is solvable via secure out-of-band channels, and where 
participants value mathematical guarantees over operational convenience.

Your point about the transmission paradox has actually inspired a potential future direction 
for GhostLine: using an initial large in-person OTP distribution to bootstrap a ratcheting 
mechanism where each session generates a new OTP for the next session. 
This would maintain information-theoretic guarantees while reducing the operational burden 
to a single initial distribution - essentially key ratcheting with information-theoretic 
security rather than computational.

This is currently just a conceptual idea rather than an implemented feature, but it directly 
addresses the operational challenges you highlighted. I would genuinely value your thoughts 
on this approach, given your deep experience with practical cryptography.

The current system explicitly acknowledges in section 3.1 that security ultimately reduces 
to the initial distribution mechanism, but once established, provides guarantees no 
computational cipher can match. Your credit card encryption story illustrates exactly why 
this matters when data size permits.

Thanks for adding such valuable historical context and for stimulating new thinking on this 
challenging problem.

Best,
Hitokiri
-----BEGIN PGP SIGNATURE-----
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=1jGW
-----END PGP SIGNATURE-----


More information about the cryptography mailing list