Re: [MUTE] Supernodes in Ants & Mute
Status: Beta
Brought to you by:
jcr13
From: closedshop <clo...@gm...> - 2005-02-28 08:42:26
|
Psychofreak wrote: > closedshop wrote: > >> >> found this: >> http://www.zeropaid.com/bbs/attachment.php?attachmentid=11074 >> > STOP LYING ! you didn't "find" this, you've posted it yourself! long version: Found this reference to the mentioned issue. Tsafa is rigth, we should think about a supernode structure for manet based protocols. This Architecture, as re-design of the "Trelawney-protocol" mixes ID-2-ID Transfers with IP-2-IP transfers. The G2 protocol with Hubs (Supernodes) is used. Tsafa was told, that is is not a mixture of 2. and 3. generation NETWORK, but tu use an OWN G2 PROTOCOL (so not the shareaza network, which was implecated by trelawney protocol design for a hybrid app). This is not connected to any second generation media filesharing network or app. The first 1-2 hops of ants/mute are done with ants protocol. The last 1-2 hops are done with ants/mute protocol as well. In the middle all hops are done by g2 protocol with supernodes structure and it is clear, that not client shared media is stnt to a supernode, but ONLY REMOTE shared media. So all is pretty safe, you have 1-2 hops plus 1-2 hops and then the long distance flight to the ants hill /colony with the g2 protocol and with supernodes/hubs. This would work pretty good. within 5 hops at maximum you can transfer the chunk and find in 2-4 hubs the whole media and can search like morpheus neo net DHT the horizont of 6 Million users in a short time, and find everything. It also provides a new gnutella - gnutellants or gmuteella - and is the future. So just update gnutella with this structured protocol, that 1-2 hops at the beginning and 1-2 hops at the end of the proxy chain allis done by a selfproxying network and the rest by worlds famous G2 network. :-) |