On a first scan read, I'm not sure there's much of interest here. As he notes in the efficiency section, it's comparable in verification cost to pre existing C(LSAG), at least that's intuitive, so I just believe it. It *is* very compact, but I don't think that's new.
I think curve trees + signature of knowledge of the rerandomised key gives the same effect with much faster verification, at the cost of a fairly trivial increase in size of proof (and with very sublinear scaling, imagine 2.6kB for 50K keys and 2.6kB for 200K keys).
In a blockchain proof size is paramount, but verification speed is also important. Luke Parker's FCMP stuff is actually trying to get both together.
waxwing /
npub1va…knuu7
2024-07-12 15:00:26
in reply to nevent1q…r0mf
Author Public Key
npub1vadcfln4ugt2h9ruwsuwu5vu5am4xaka7pw6m7axy79aqyhp6u5q9knuu7Published at
2024-07-12 15:00:26Event JSON
{
"id": "d9d382435220a22ae16dc73ce36df6e67f7b2d9fe81affdef932daff34f30ca3",
"pubkey": "675b84fe75e216ab947c7438ee519ca7775376ddf05dadfba6278bd012e1d728",
"created_at": 1720796426,
"kind": 1,
"tags": [
[
"e",
"f2a4bacd9f9bee2198a7c15f485b0d9c9244a8d644249bdaf917c18cf60e1076",
"",
"root"
],
[
"e",
"0a24e60601dcb524fa022c4192bc7e7c497a09fd571a301dc953885416c3667e",
"",
"reply"
],
[
"p",
"675b84fe75e216ab947c7438ee519ca7775376ddf05dadfba6278bd012e1d728"
],
[
"p",
"e0cfb5549d3cf7db4e2736f8e1bc84f62486af7a41295d867c6a313459042528"
],
[
"r",
"2.6kB"
],
[
"r",
"2.6kB"
]
],
"content": "On a first scan read, I'm not sure there's much of interest here. As he notes in the efficiency section, it's comparable in verification cost to pre existing C(LSAG), at least that's intuitive, so I just believe it. It *is* very compact, but I don't think that's new.\n\nI think curve trees + signature of knowledge of the rerandomised key gives the same effect with much faster verification, at the cost of a fairly trivial increase in size of proof (and with very sublinear scaling, imagine 2.6kB for 50K keys and 2.6kB for 200K keys).\n\nIn a blockchain proof size is paramount, but verification speed is also important. Luke Parker's FCMP stuff is actually trying to get both together.",
"sig": "6978a5d41f0d8d455c7ba40579b8432e3214530b9e9f3a77e7bcf72dec88b83c9cf5715502fc0735be54b9866350c83a2b67677b7f7fae1565861fd63c66ba52"
}