Mr.Bad (Tue, Jun 19, 2001 at 10:06:19AM -0700):
>     m> you can't force a node to limit the references/IP
> Why not? Please elaborate.

well, YOU can force YOUR node, but freenet-nodes can't be forced in
general.  just remove the part of code.  (if added, there most probable
will be a option for turning out this feature)

>     m> two peers are not very much...  i wouldn't consider freenet
>     m> anonymous with 2 peers per node
> I don't think you're addressing my question. As was mentioned before,
> "young" nodes often settle to only 1 or 2 node addresses in the data
> store. I don't like it, you don't like it, nobody likes it. How do we
> fix it?

want to say:  with only two peer-nodes, you have other problems.

initially... well... 



>     m> (which, BTW, can't be a network)
> 
> What? "Can't be a network"? Is that for some weird definition of
> "network" where every node has to be connected to 3 or more other
> nodes?

well... for me, a network with 2 peers/node is a ring.  in such
a network, thinking much about routing would be a waste of time.

>     m> that's a client which is caching downloaded data.
> 
> Let me be more specific, since you're insisting on being some kind of
> literalist fuckhead.
> 
> Let's say node has 10 addresses that it glooped from inform.php. Over
> time, one of these addresses begins to predominate, and the other
> addresses stop responding to Freenet requests. Soon, there's only one
> address still alive. What do we do then?

send a request to that node, get the reply, and, with a high probability
(70% of the requests that big node doesn't handle (given the nodeoperator
doesn't change the probability for resetting dataSource)), you get the address 
of
an other node.

> Redefine the node as not-a-node? That's your answer?

Thought you artifically limit the peers to one node, as in your example
above.  well, i've never heard about a node loosing all but one reference,
but thaen, again, not limiting the references/node doesn't help, then
the protocol is fucked.

-- 
moritz

.------------------------------+------------------------+----------------------.
|    mailto:moritz at 1723.net    |    http://1723.net/    |    ICQ: 110110488  
  |
|------------------------------+------------------------+----------------------|
| this  mail is  irevocably placed  in the  public domain  and  may be  read,  |
| redistributed    and   erased    freely,    now    and    forever,    amen.  |
?------------------------------------------------------------------------------?

_______________________________________________
Devl mailing list
Devl at freenetproject.org
http://lists.freenetproject.org/mailman/listinfo/devl

Reply via email to