Post A2LJau4DSkToPXkT4K by leip4Ier@infosec.exchange
 (DIR) More posts by leip4Ier@infosec.exchange
 (DIR) Post #A2LJZ8KjRwKa4ud6XY by leip4Ier@infosec.exchange
       2020-12-19T00:54:57Z
       
       0 likes, 0 repeats
       
       i'm not sure it'd work in real life, but what if?a distributed social network in which each user has a set of signing keys. one key is used to sign public posts, another is for private ones. the user also signs the list of people who can see private posts. posts are announced to the network and then copied by other peers. but, if someone shares a private post to someone who shouldn't have access, software of that user generates a proof that it received a post it shouldn't've and broadcasts it.
       
 (DIR) Post #A2LJau4DSkToPXkT4K by leip4Ier@infosec.exchange
       2020-12-19T00:55:14Z
       
       0 likes, 0 repeats
       
       then everyone who received that proof block the bad person, effectively banning them from the network. same for distributing deleted posts.it could use maybe a bloom filter instead of a list of allowed users. and also this system would require all network interactions to be signed.
       
 (DIR) Post #A2LJlSeJk1x8YtfX8a by leip4Ier@infosec.exchange
       2020-12-19T00:57:17Z
       
       0 likes, 0 repeats
       
       (i'm not sure the problem i'm trying to solve existed in the first place, it's just sleepythoughts, i don't know if they make sense)
       
 (DIR) Post #A2LK4liH5RM3Bnj5zU by varx@cybre.space
       2020-12-19T01:00:00Z
       
       0 likes, 0 repeats
       
       @leip4Ier You'd run into issues with distributed event ordering. If I make a post, you share it, and then I delete it... other people might see it as "you shared a deleted post".Plausibly the same could happen if I change the visibility on a post (if that's even supported).Could be weaponized against other users.
       
 (DIR) Post #A2LKKaYyJHo3F1v1Bg by leip4Ier@infosec.exchange
       2020-12-19T01:03:31Z
       
       0 likes, 0 repeats
       
       @varx isn't there a way to reliably synchronize time? i assumed that any broadcasts are silently dropped if the time difference is more than like 10s.
       
 (DIR) Post #A2LMR52S7EKBPBvnJQ by varx@infosec.exchange
       2020-12-19T01:26:53Z
       
       0 likes, 0 repeats
       
       @leip4Ier I don't think this is a case where you can just trust everyone's timestamps (maybe I'm wrong, haven't thought it through carefully) and in any case people's computer clocks are frequently mis-synced anyhow. I suspect 10s is going to be too tight for a distributed system anyhow -- without the possibility of asynchronous delivery, you're likely committing to a certain degree of centralization.
       
 (DIR) Post #A2MnKdxklHpLc8qehM by ketmorco@mastodon.technology
       2020-12-19T18:03:05Z
       
       0 likes, 0 repeats
       
       @leip4Ier that would be fun to write. I mean, the software would be spoofable, but.