On Wed, Feb 06, 2002 at 04:32:12AM -0600, Mark J Roberts wrote:
<> 
> Am I the only one who has these preposterous court arguments
> shredding up my brains? Like this one:

The goal is to make Freenet stable under attack - clearly if people can
censor data completely by removing the key from their own node (or a
couple of nodes) then the concept isn't working. But there is clearly
also a limit to the amount of bad behavior that Freenet can hold up to -
a single node routing completely wrong won't hurt the network, but if
50% of the nodes started doing it that would be a problem. 

There are many things that a node can do to try hurt the network, our
hope (/wish) is that the network will be able to hold up against those
who would do this.  However, since there is inevitably a limit to how
many attackers can be handled, it does make sense for us not to include
and thus encourage functions that are hurtful to the network. It is
possible to flood the network, but I think we'll all agree that a
DoFlood FCP comment is a little unnecessary. I don't really see removing
data as an action in the spirit of what we are trying to achieve, and
beyond that I worry about confusing users (who might think that it
actually deletes things from Freenet), so I have a hard time convincing
myself why we should have such a function.

On the other hand, all the silly "but my eyes were closed when I fired
the gun" arguments are most definitely stupid. If something is possible,
then it has to be considered as good as implemented, both for users and 
attackers (oh yeah, Gianni, you don't believe that...)

-- 

Oskar Sandberg
oskar at freenetproject.org

_______________________________________________
Devl mailing list
Devl at freenetproject.org
http://lists.freenetproject.org/mailman/listinfo/devl

Reply via email to