💾 Archived View for midnight.pub › replies › 4747 captured on 2023-03-20 at 21:02:36. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2022-07-16)
-=-=-=-=-=-=-
< The Spring of '83 is not over
Okay, I have, now, actually read the spec. There is a very narrow core of it that I actually quite like (keypair identities making authentication of posters and validation of posts trivial), and I like the general user experience goal... But there's a whole lot of other stuff I don't like. I do not like any "boil the oceans to prevent spam" mechanism, and this protocol has two (related ones) of them: the need for publishers to search the keyspace for keypairs that meet the protocol's requirements, and the "difficulty factor" that servers must apply before deciding to accept posts. I don't know *exactly* how much compute it would take to generate a key that is both compliant and high-difficulty, but I do know it provides bad incentives: rather than disincentivising spam (by making it expensive), it makes *only* profitable material postable by making the protocol pay-to-play once it's widely used.
I also don't particularly care for the 'realm' system. It makes sense, once you've accepted some of the other design goals (particularly "store the whole universe"). But there's not a huge benefit to accepting those goals to begin with.
Yeah, I'm similarly torn on the spam prevention mechanism, but realistically, I can't think of any relevant prior art to point at as proof that it either works or it doesn't, so having thought about it for all of five minutes, it's more likely that the author is right and it'll work than that I'm right and it won't.