I've been thinking about alternatives to digital systems that assign users "roles" with varying "privileges" and as a result (unintentionally) enforce a hierarchy between users. I also don't want to encode too much theory/procedure into a digital system, because that just makes it rigid and brittle -- procedure should be in the realm of human beings working together. I think I have something that's simple and interesting, at least interesting enough to try.
most systems have what's called "access control lists" which are basically records that look roughly like
(resource:id, user:id, read:bool, write:bool, delete:bool)
so given a resource and a user, what are that user's privileges with respect to the resource.
I'm thinking more along the lines of consensus among good faith actors. the system is invite only and the default is that anyone can do anything on their own. if certain resources require more vetting the community can increase the size of the consensus required to take certain actions.
in place of an ACL we have records like
(resource:id, write:int, delete:int)
so, given a resource a certain number of community members must agree on a write or a delete before it goes through, default being 1
@nasser is that different from blockchain consensus?
@ivan yeah this is just saying a number of people in the community have to agree to take an action, unrelated to blockchain consensus
@nasser ok just really sounds like blockchain consensus :)
@ivan I don't see it, can you point out the similarities?
proof of work blockchain consensus as I understand it requires one actor to build a longer chain than anyone else and propagate that through the system, getting everyone to agree that that is the new head of the chain.
in the system that I described there isn't necessarily a chain, nor the need for global consensus, nor pressure on a single actor to outcompete others.
@nasser it seems similar because blockchain consensus is also people reaching a threshold consensus on changes to a database. it's just adversarial because of the public nature of the ledger, which solves the central authority issue you were mentioning and proof of work is just one approach.
also not sure what you mean by 'pressure on a single actor to outcompete others' and why 'no need for global consensus' in what you're thinking. isn't there always some sort of quorum threshold?
