Oskar writes:
I have been saying we should decrease this to 4-5 hops on
average so as to get a slightly more stable namespace, which I
think will give it better characteristics...
I was thinking about that too, but here's another theory:
Setting the DataSource to your own node address doesn't really help
anyone; the recipient obviously knows about you -- after all, they're
the ones asking you for the data. What the recipient really needs to
know is where *you* got the data.
So, my proposal is that a node should, with fairly high probability,
set the data source for the message to the ID of the node that sent the
current node the data.
Now, this gets problematic for the original data source for the node. I
therefore propose that nodes should, as an alternative to claiming the
source of the data they are sending, occasionally delete the data source
field all together. That looks better from the standpoint of anonymity,
too -- instead of falsely claiming "yeah, I'm the source for the data,"
nodes can either claim "yeah, them's the source for the data" or "the
source of the data is none of your business." If you're going to lie
about where you got the data, better to blame it on someone else. (-:
_______________________________________________
Freenet-dev mailing list
Freenet-dev at lists.sourceforge.net
http://lists.sourceforge.net/mailman/listinfo/freenet-dev
