On Sunday 10 April 2011 19:25:53 Volodya wrote: > On 04/10/2011 08:56 PM, xor wrote: > > On Sunday 10 April 2011 10:57:45 Volodya wrote: > >> I am likely missing something trivial here, but hopefully the community > >> will find what i wrote interesting. > > > > People could just create single-use identities instead, no code changes > > would be needed, you can do this right now already. > > > > If many start doing it we could add a "valid-until-date" to identities to > > optimize the fetching process to not fetch content from single-use > > identities after they are obsolete. > > Yes, of course people could do that. But that would cause 3 things: > 1. The messages of announced, but non-established identities would need to > be more visible than they currently are.
We cannot make them any more visible than that without allowing DoS. > 2. People would need to predict in advance the need for an extra identity, > and keep solving puzzles with it. When you realize that you need a new identity you can create it immediately. You can solve the puzzles after having created it. This would work the same way with anonymous identities. > 3. A lot more pulling/subscribing going on. Which makes the whole thing > even less likely to survive large scale use. Can be avoided by adding a valid-until-date. > On frost i have uses several identities, i would usually post as anonymous > about once every 3-4 days (before the DDoS stopped almost all the > conversations) and i don't think i was an exception, in fact some may have > done more. This would mean that 2 new identities would have to be created, > and announced every week. I don't remember the last estimate of the size > of OpenNet, but lets say its 2000 people, that's 4000 *throw-away* > identities every week if all people start using FreeTalk actively. It doesn't really change with your proposed "anonymous identities": IF we allow "anonymous identities" then we need to make sure that their posts - if spam - can be removed. This means that there must be some way of separating them in the database - which results in an IDENTITY being created for them which can be separately distrusted. Actually the semantics of your "anonymous identities" are equal to what we have now in such a way that it would result in the same implementation = no code changes at all EXCEPT for a valid until date. Think about how to program this if you're able to, you'll realize that the implementation is what we currently have... -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 198 bytes Desc: This is a digitally signed message part. URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20110422/459c4957/attachment.pgp>
