IRC channel logs
2023-04-18.log
back to list of logs
<fr33domlover>jfred: The current model I have is based on the AP fediverse, i.e. Mastodon etc., with user@host identities. It allows to blacklist users (so the public endpoint would check you against that blacklist, seems reasonable?). Indeed creating new users is easy, and if a server generates a lot of abusing users you then blacklist the whole server. Now the <fr33domlover>question is, can someone cheaply mass-generate domain names and host software to communicate via those domains. AFAICT this hasn't been a challenge on the Fediverse, maybe because mass-generating domains like that isn't cheap/easy enough even for dedicated spammers <fr33domlover>I guess in a more P2Pish setup, some sort of WoT would be needed <sneek>Welcome back Zarutian_iPad!! <Zarutian_iPad>I am also thinking about Liquid Democracy in this context as in trust-revelancy <Zarutian_iPad>plus the “float-switches” in the “leaky bucket/integrator” from subsumption architecture <Zarutian_iPad>in the context of wiki/google-wave-esque-thing/blog/MUD-or-MOO this can be used as a linear-numerical input to calculate cost/postage that spefic entity must pay/expend to perform a spefic action or edit <Zarutian_iPad>the cost could be in time, that is limit on how often the action can be performed, et ceterata <Zarutian_iPad>this could also be used for n-of-k kind of situations (like in cybercoinz multisig) involving any kind of “access structure” (term from cryptography literature) to perform certain actions <Zarutian_iPad>ACTION notes the idea of combing use of access structures with deniable encryption plus say something like the ink DSL for need to know basis documents. Lower clearence meaning more redacted and in certain ways.