Hello everybody,
I'm interested in P2P networks and more precisely in trust relationship. I have read some papers about Kad but a point remains dark for me concerning the different types of ID used in KAD and eMule.
For instance in
http://www.cs.huji.ac.il/labs/danss/presentations/emule.pdf, page 8 , it's written that there are 2 different ID in eMule. The first one called userID is used by the credit system, identify the "user / computer" using eMule. This userID doesn't change between sessions. Another ID is presented called clientID which is only valid through a client and a server for a specific session. So eMule has these two kinds of IDs.
Now concerning Kad,
http://www.eurecom.fr/~btroup/BThesis/Mast...esisBrunner.pdf page 36, the Kad clientID (KadID ?) which is used for the routing in Kad network is described as an ID that doesn't change across sessions. I would like to know if this description is true because usually in structured P2P networks the clientID used for routing changes between sessions. If the clientID doesn't change in KAD and isn't redistributed after a disconnection, I don't see how the kademlia tree can evolve properly.
Concerning the credit system in Kad, I would like to know which ID is used. Is it the clientID used for routing or the userId described in eMule.
If somebody can bring me light on these points, I would be grateful.
Thank you.
