[BitTorrent] Re: Back to Merkle Hash Trees...

Joseph Ashwood ashwood at msn.com
Mon Feb 7 19:31:17 EST 2005



----- Original Message ----- 
From: "Olaf van der Spek" <OvdSpek at LIACS.NL>
Subject: Re: [BitTorrent] Re: Back to Merkle Hash Trees...


> Which definition of merkle trees are you using?

My assumptions except where otherwise noted have been that the branching is 
arbitrary and determined at execution time. Specifically, the tree could be 
root with 2^32 direct children, or root with 2 children, each with two 
children . . . . . 32 levels later the leaf nodes, or any other useful 
configuration. This allows the majority of useful behavior to flatten the 
tree completely (depth >2 hash trees aren't really useful for less than a 
few thousand nodes, and arguably not useful until millions have been 
reached) while the situations that will benefit from certain behavior in 
Merkle trees can create the depth which will be of most benefit. My examples 
were limited to 3 nodes only because I wanted to avoid too much typing. I'm 
not sure if I said it or not, but it is my opinion that the number of nodes 
variable should be at least 16 bits, preferably 32, although 64 is 
distinctly overkill since 2^64 256K blocks at 500Mbytes/cubic mm density 
would occupy 0.9 cubic km.

In my design I also assume that the branching may not be known until the 
node is reached, so the tree may actually have dynamic branching where nodeA 
may have 2 children, nodeB 17, and nodeC 482.
                Joe 






 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/BitTorrent/

<*> To unsubscribe from this group, send an email to:
    BitTorrent-unsubscribe at yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/
 





More information about the BitTorrent mailing list