On Thu, Jul 19, 2012 at 2:07 AM, Michael Rogers <[email protected]>wrote:

> So, if I understand right, both parties must sign off on a transaction
> in order for a third party to include the transaction in its
> similarity calculations? So Sybils can create fake transactions
> between themselves, but not between a Sybil and a legitimate peer?
>

Confirm. This can also be used as a mechanism for 3rd parties to spot when
relationships between other peers broke down (i.e. there are disagreements
about history in the long chain certificate), so peers analyzing this data
set can look for patterns of abuse among multiple peers.


> If that's right, I only have two remaining concerns: bootstrapping and
> positive feedback. The problem with bootstrapping is how to
> distinguish legitimate peers from Sybils when you have no interaction
> history.
>

Bootstrapping is a hard problem in any file sharing system including
torrent trackers.

I'd like to use social features to help with the bootstrapping problem. I'd
like for peers to be able to identify their "friends" which would enable a
sort of free-leech mode. "Friends" could either be several peers controlled
by one person, or a Tahoe Friendgrid-style system where the participants
generally trust each other.

To bootstrap a new peer, you can request popular content from one of your
friends, and serve it to the rest of the network. Your friends can then
indicate you have the content available and direct other peers to obtain it
from you.


> The problem with positive feedback is how to prevent the interaction
> graph from collapsing into clusters or isolated components.
> Interaction leads to similarity, similarity leads to interaction -
> don't peers become more and more likely to interact with the same
> peers? Is there a need for a countervailing mechanism to encourage
> exploration and give new peers a chance to interact?


I'm not sure clustering is bad. I think ideally the system clusters around
groups that are both interested in the same content and have a favorable
network topology between them.

-- 
Tony Arcieri
_______________________________________________
p2p-hackers mailing list
[email protected]
http://lists.zooko.com/mailman/listinfo/p2p-hackers

Reply via email to