On Friday 18 January 2008 12:42, Michael Rogers wrote:
> On Jan 18 2008, Matthew Toseland wrote:
> > All the nice maths about weighted coins assumes when we throw the failure 
> > we kill the request. On this basis, a 10% pDrop equals an average 10 hops 
> > path length.
> 
> Right.
> 
> > You are proposing to auto-retry on a DNF, presumably to prevent short 
> > paths. This changes the average path length for a 10% pDrop to something 
> > on the order of average node degree * 10 i.e. around 150-200.
> 
> Depends what you call the path length, I guess - the distance from the root 
> to any leaf of the search tree is 10 hops on average, but the search tree 
> can have many branches. If you want 10 nodes to be searched on average then 
> we should flip the coin again on a DNF.

Eh? If you want 10 nodes to be searched on average then you need to kill the 
request on a DNF, haven't I adequately demonstrated that? If you retry, even 
if you flip a coin deciding whether to retry, then you greatly increase the 
average number of nodes visited.
> 
> But this is all moot because weighted coins seem to require tunnels or 
> premix routing. :-/
> 
> Cheers,
> Michael
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: 
<https://emu.freenetproject.org/pipermail/devl/attachments/20080118/809d9185/attachment.pgp>

Reply via email to