for the routing process, 如果找不到，应该发到最远得邻居，可目前得问题是怎么保存最远得邻居.我想了一个办法就是把query加上一个经过得hop list，比如 query was sent by A, forward to B, and then to C.... until it arrives X, say it has 10 hops in between, then X knows A is 10 hops away from it. So it add A as a 10 hops away neighbour. similarly, it can keep some 5, or 8 hops away neighbour. Assume the current radius is 5, When it cannot find match in its routing table. it can should forward the query to its 8 or 10 hops away neighbour, because the 1-5 hops neighbour for sure don't have the match. so it don't need waste the time. How do you think of it? Also this way, can avoid a query forward a node already be forwarded.就是碰到了，就记下来是 记下来也可能变,那当然了，不过还可以更新啊，下次又有一个query过来，再记下一个,总记最新得
