Hi,
Georg Koppen:
I think I am confused a bit. So, how does that relate to the contact information sharing specification you propagate? Is your new proposal an additional thing relay operators should implement on top of the that specification? Or should they choose between the two? What shortcoming does your new proposal solve that is not addressed by the other specification and vice versa?
On a technical level CIISS proofs [1] and HAROI proofs are the same, the main difference is the integration into tor and the verification of proofs by directory authorities.
The proof field in CIISS would eventually become obsolete should HAROIs get implemented in tor, but since the proof is the same, relay operators do not have to setup some new kind of proofs when HAROI is implemented (>1400 relays, >50% exit probability have properly setup their proof already and more will follow soon). The CIISS proof will continue to serve its purpose until HAROI is deployed in tor releases since it naturally takes a long time until all relays run a supported tor version that would support it.
The main benefit of HAROI is the central verification of proofs by directory authorities instead of requiring everyone to verify the proofs themselves. This is better for efficiency and will reduce the load on proof endpoints (DNS and webservers).
I hope that helps clarifying the relation between HAROI and CIISS proof field.
Should you have any more questions do not hesitate to ask.
kind regards, nusenu
[1] https://nusenu.github.io/ContactInfo-Information-Sharing-Specification/#proo...