> Oh no. Freenet *relies* on path compression. Without it you can't use a
> linear search to get data. A couple of studies have been done on
> small-world networks like Freenet, where its shown that without far
> reaching links (created by path compression) they fail. One is included
> in an upcoming orielly book on P2P.
Yes, I've read it.
> Creating a cluster means that you
> have to increase the overall htl by the size of the cluster in order for
> requests into the cluster to work.
Not at all. That would mean that you searched the entire cluster first,
which is ridiculous and not how Freenet routing works at all. A cluster
would be a fully connected graph, meaning that you are one hop from the
gateway. You would follow the best path through the cluster just like
normal.
> Now, imagine three clusters (just three little clusters) of maximum width
> 10 exist on Freenet, and a request hits all three of them. With HTL 15,
> each gateway has to increase the HTL by 10 to make it appear as one node
> (as you so fondly suggest a cluster is). That means that you have an
> effective HTL of 45 for this request. Is it just me, or is that an
> unreasonably long time to wait for data?
I can see why you dislike clusters so much. But that's not how it would
work at all.
So let's say that 99% of Freenet users don't use clusters, as
you've suggested. Despite that, you still happen to hit 3 gateways in a
row! Oh my! The first gateway, we assume, does not have the file. It finds
the next best node. Assuming that this node is one in the cluster (it's a
normal public node as well as being a gateway and so knows of other public
nodes), it can simply forge the HTL to be 1 before requesting from the
best cluster node. Assuming the request times out, it decrements the
original HTL and passes the message on.
So 3 clusters, assuming that every time there is a gateway the the next
best node is in the cluster, will take 6 HTL. Each gateway node takes 1
HTL and each cluster takes 1 HTL. This is the same as if you just had the
same situation happening with 6 nodes.
> Media-enforcer attacks are closed with P/K and varying htl decrements. So
> quit harping up on that. Hostile-territory Freenets need to deal with
> this problem more thoroughly, with solutions outside of Freenet
> (super-paranoia stuff like stego).
Except that MediaEnforcer attacks aren't stopped with varying htl
decrements or else I wouldn't be bringing this up at all. PK doesn't stop
them, it just means it takes longer to get IPs. The only thing that will
stop them is to reject connections from unknown hosts. PK definitely does
make it more difficult, which is great, and I very much look forward to it
in the future. Varying HTL decrements, though, doesn't change anything.
> > which probably won't happen. However, even so, a slight tweaking is
> > probably possible to eliminate this problem entirely.
> Which would only work for small clusters, since you should need more than
> one hop to find data in any non-tiny cluster. (Chinese dissidents are
> going to have more than a few nodes).
It's not as black and white as that. First of all, there wouldn't be a
single gateway between China and the rest of the world. But if there were,
you wouldn't need tweaking. If you have a million nodes in a cluster then
there is no reason to treat it differently than a normal network. So there
is no point in making the cluster act like a single node. The country of
China can absorb as much HTL as it wants if it has a million nodes.
HTL tweaking is only necessary or useful for small clusters.
> > Except if you can't tie a user to an IP address, of course, but you can in
> > many schemes. For instance if you keep logs of your DHCP transactions then
> > it's not a big deal, even on cable.
>
> Having a cluster wouldn't really help you. They'd find the gateway, kill
> the leader, but only after torturing him to get his nodes.config for the
> list of the other nodes in his cell. You need more than just this
> clusters hack to prevent it.
This is a problem with trust webs. If you trust people they can betray
you. This is still a heck of a lot better than trusting everyone. In
which case they'd just find you all and kill you all.
_______________________________________________
Freenet-dev mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/mailman/listinfo/freenet-dev
