On Tue, 16 Apr 2002, Mark J Roberts wrote: > Kevin Atkinson: > > The probability of selecting three malicious, buggy hosts will be very low > > since part of the selection criteria will be on past performance. > > Furthermore it will likely select the same couple of hosts to download the > > blocks of the file. > > Biasing node selection based on past performance is probably > illogical. If a reliable node emerged on to the network, it would be > recognized as reliable and flooded. Then it wouldn't be reliable any > more.
I just have to see. There are things I can do to avoid this problem. > I'm still unconvinced that block retrieval is as reliable as you say > it is. Even if a transient error or malicious host can never cause a > block failure, the network's algorithm itself may not be reliable. It may be slightly less reliable than retrieving data as a whole but I don't think it will be as unreliable as you think it is. I think we will just have to agree to disagree. > Is it reliable? (where reliable approaches 1) Is freenet? It sure doesn't seam that way. I am convinced it can be made at least as unreliable as freenet. --- http://kevin.atkinson.dhs.org _______________________________________________ freenet-tech mailing list [EMAIL PROTECTED] http://lists.freenetproject.org/mailman/listinfo/tech
