On Thursday 07 May 2009 11:09:11 Ximin Luo wrote: > xor wrote: ... > > >> On a different issue, I don't think it's a good idea for the algorithm to > >> recurse indefinitely. Would you be able to make it automatically construct > >> a small world network? Download the trust list of your most trusted IDs, > >> then (say) 2/3 of their most trusted, then 1/2 of theirs, etc? > > > > It is absolutely necessary that the algorithm downloads ALL identities because > > with messaging systems like Freetalk ALL messages should be visible to the > > user, not just "some" messages. > > we will have to come up with a way to *make* this work because contacting > everyone in a network cannot (as in mathematically impossible to) scale.
More is possible than you might at first assume, in the medium to long term. A network of subscribers can be very efficient, assuming ULPRs and full blown passive requests, provided that subscribers are waiting and propagating rather than constantly polling. I am assuming that the number of actual messages posted is acceptable; this may or may not be a good assumption depending on how much segregation there is in the community between users of different boards. However, I have always assumed that there would be some higher level propagation hinting behaviour. When a message gains a user some trust, the message itself should be propagated if possible. IIRC there is some amount of this on the current system. We can probably improve on it further. > > i don't see why it is necessary to do as you say. here is my understanding of > how freetalk works: > > - gets a list of IDs with good trusts > - polls the outboxes of them > - downloads and displays new messages to the user > > in a web of trust, each node clearly does *not* think everyone as equal. i > don't see this to be a problem or hindrance to liberty. in fact it would be > pretty authoritarian if everyone had to treat everyone exactly equally. > > in the case of enforcing free-speech, if there are more censor IDs in a network > than free-speech IDs, then the former will win unless you arrange things > hierarchically such that the free-speech IDs are on top. then (as in any > hierarchy) you will have the problem of judging who is "free-speech". > > ultimately, a WoT must reflect the wishes of its population; nothing more or > less. **we cannot go against the wishes of the population**, we can only make > it such that, **if** most people desire free speech, **then** they are capable > of having it. we can only take a horse to water, etc. > > that said, it is easy to have freedom of speech, in the sense of being able to > reach people that want to hear you. in the case of releasing top-secret > documents, you can have something like this setup: > > - one ID acts as a 3rd-party publisher, that everyone trusts (cf. wikileaks) > - the informant gets trust in this publisher > - the informant publishes their dox > - the publisher polls and retrieves the dox and in turn publishes it in their > outbox > - everyone downloads it > > this setup can also be used to "enforce" freedom of speech - set up publishers > that agree to echo all non-spam messages that get sent to them. > > obviously this adds some degree of centralisation in the system, but you can > re-decentralise by having several publishers that trust each other etc, sort of > like Usenet. when one publisher is compromised, the others can then de-trust > them. etc etc etc. > > anyway, sorry if all of this has been discussed and said before. i'm new to all > of this. feel free to tell me to RTFM (but also tell me where TFM is ;) > > X -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 835 bytes Desc: This is a digitally signed message part. URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20090507/b972c672/attachment.pgp>
