--- Day changed Mon Oct 28 2019
01:57 -!- viasil [~viasil@95.174.67.204] has quit [Ping timeout: 265 seconds]     
01:59 -!- viasil [~viasil@95.174.67.204] has joined #joinmarket     
02:38 -!- Zenton [~user@unaffiliated/vicenteh] has joined #joinmarket     
03:14 < waxwing> belcher_, did you see this? slightly strange, it's very short and seems mostly just inspired by your recent fidelity bonds discussion: https://eprint.iacr.org/2019/1111.pdf
06:44 < belcher_> hah nice
06:45 < belcher_> ty for the link, how did you come across it?
06:51 < belcher_> gmaxwell pointed out on reddit a few days ago that monero can also be targetteed by sybil attackers, that would work by an adversary spamming the chain with lots of their own outputs, since users randomly choose outputs to include in their ring signature if a user chooses only the spammers outputs then the spammer can know the real user's inputs by ellimination
06:52 < belcher_> so monero's sybil resistance depends on its miner fees
06:56 < waxwing> oh sure the model is very similar, it's just stronger mainly because of the ability to do it without coordination.
06:56 < waxwing> and anyway, nowadays CT, so .. etc.
06:57 < waxwing> re: paper, found it on bitdevs.org latest link list (i mentioned a couple days back, that site)
06:59 < belcher_> ah     
07:05 < belcher_> wow they also missed a big point about the fidelity bonds
07:06 < belcher_> they write "The assumption is that the adversary wouldn’t want to Sybil attack a mixer which would hurt the long-term value of their holdings through making the coin less private." which is wrong
07:06 < belcher_> thats not "the" assumption, its just an interesting side-effect which it seems many people have latched onto
07:07 < waxwing> maybe i'm too dismissive but that's dumb because that's just an argument that nobody would ever attack bitcoin in any way, ever (not a very safe assumption!).
07:08 < belcher_> the important part of the whole scheme is that a fidelity bond value goes as the square of the coin amount, that means that if a sybil attacker wants to create multiple identities they face a quadratic barrier, i.e. if they have V coins and split them up amongst 5 sybils then they've actually reduced their fidelity bond values by 25 (= 5^2)
07:08 < belcher_> the V^2 term is the crucial assumption, it gives a disproportionate advantage to the defenders of privacy
07:09 < belcher_> and yes as you say the other assumption is not very strong (in hindsight i shouldnt have even put it to paper)
07:21 -!- belcher [~belcher@unaffiliated/belcher] has joined #joinmarket     
11:47 -!- Zenton [~user@unaffiliated/vicenteh] has quit [Ping timeout: 250 seconds]     
11:54 < waxwing> hmm yes it's often something mentioned, and it's not unreasonable depending on context, but it's probably best to leave it out of arguments generally.
11:57 -!- reallll [~belcher@unaffiliated/belcher] has joined #joinmarket     
12:01 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 268 seconds]     
12:14 -!- reallll is now known as belcher     
13:28 -!- belcher [~belcher@unaffiliated/belcher] has quit [Quit: Leaving]     
13:44 -!- Zenton [~user@unaffiliated/vicenteh] has joined #joinmarket     
14:40 -!- MaxSan [~four@195.206.105.227] has quit [Quit: Leaving.]     
16:57 -!- nsh [~lol@wikipedia/nsh] has quit [Max SendQ exceeded]     
17:15 -!- nsh [~lol@wikipedia/nsh] has joined #joinmarket     
18:48 -!- CgRelayBot [~CgRelayBo@p54866B19.dip0.t-ipconnect.de] has quit [Ping timeout: 246 seconds]     
18:49 -!- AgoraRelay [~jmrelayfn@p54866B19.dip0.t-ipconnect.de] has quit [Ping timeout: 276 seconds]     
18:59 -!- CgRelayBot [~CgRelayBo@p5DE4A37C.dip0.t-ipconnect.de] has joined #joinmarket     
19:05 -!- AgoraRelay [~jmrelayfn@p5DE4A37C.dip0.t-ipconnect.de] has joined #joinmarket