QRT based node selection

#### This is an idea, and far from draft state ####

Do you think node selection based on similarity in the QRT would be useful? 


Reasoning for doing so: 

* People with similar QRT table share stuff similar to the stuff I share 
myself. 

* People who share stuff similar to what I share are more likely to have stuff 
I search for right now. 

* QRTs are exchanged and compared already, so nothing needs to change in the 
Gnutella protocol. 

* Node selection based on similarities in the QRT would group people with 
similar interests together. 

* That way the first hops would be more likely to return useful query results, 
so Dynamic Querying would step in earlier and searches would consume less 
network resources, and since people often have obscure interests and share 
the stuff they are most interested in, rare files would be more affected by 
this than popular files - and spammers might group themselves together 
automatically :) . 

* And sharing stuff you like would be encouraged, since it would improve your 
search results. 


What do you think about selecting/priorizing nodes from the Host catcher based 
on similarities in their QRT? 

Best wishes, 
Arne
-- 
Unpolitisch sein
Heißt politisch sein
Ohne es zu merken. 
- Arne Babenhauserheide ( http://draketo.de )
-- Weblog: http://blog.draketo.de

-- Ein Würfel System: http://1w6.org - einfach sauberere (Rollenspiel-) Regeln

-- Mein öffentlicher Schlüssel (PGP/GnuPG): 
http://draketo.de/inhalt/ich/pubkey.txt


[Non-text portions of this message have been removed]