Matthew Toseland wrote: > On Tuesday 14 August 2007 02:20, Edgar Friendly wrote: >>>> Do you suggest that each established user create *cash puzzles, and new >>>> posters choose one or many of these puzzles to solve in order to get >>>> through untrusted/unknown-user-blocking so they can start to earn trust? >>> Yes. >> Okay, here might lie a hard part - if there's 100 established users, it >> seems impractical to solve all 100 puzzles (although I admit it could >> become necessary in order to find someone's to pay attention). How to >> choose which puzzles to solve? Should a new user really be presented >> with a list of 100 check-boxes to pick between? What information should >> they have to make an informed decision? I guess puzzle difficulty >> differentiates established users. What else? Could Frost to choose a >> bunch of people at random, or even publish more and more solutions as >> time goes on until someone evaluates the poster? > > I had thought Frost would choose some established posters at random, and use > their puzzles. Each user's Frost would present the post, and if the user did > nothing, the newbie would gain minimal trust from that user, which would be > propagated. 3 established users shouldn't be a problem
At the moment, I lean towards new users replying to posts by established users. They would have some sort of *cash puzzle included in the post, and the first reply containing a solution would bypass filtering and be presented as a user looking for credibility. The established user could then give positive or negative trust (or 0.00 trust, not some small positive amount), and publish his web of trust to a public board. > Publish it under their SSK somehow - probably publish a USK, refer to its > current sequence number in their posts. The USK could contain: > - A series of boards (encrypted so at least a dictionary attack is needed) on > which the user has recently posted. > - Current hash+think cash puzzles for each board. > - Recent posters to each board. > - Trust info (for each board?). > - Puzzles for private messages. > > Okay, maybe this is too much. That does seem too much. I propose a simple system for the small community we have now. As we outgrow it, we can learn from our successes and failures to design the next system. E.
