> I don't disagree that this is a desirable feature, but I just don't think
> that the current hysteria over it is warrented, and I certainly don't
> think that the knee-jerk reaction of implementing your proposal which I
> think is badly flawed will work.
For me this is not a reaction. I have been firmly in favor of clusters for
a long time. I think implementing it now would be good timing so as to
combat the bad reactions people will have from running that article (such
as "I can't run a node! I'll lose my Internet access!" and "Freenet is
vulnerble to attack!" and such). But of course we can't implement
something that sucks just to assuage bad press. So we must no procede by
either suggesting an alternative to my proposal which is better or to
examine the flaws in my proposal in hopes of making it passable or
figuring out how to make a better proposal. I feel that I have refuted all
flaws with my proposal up to this point, so I need some more criticism.
> crossing the road. The more people using Freenet, and the more acceptable
> running a Freenet node is, the less of a problem this is. Getting
> hysterical about "security flaws" is not helping. We should consider ways
> that we can make "node fishing" less effective, but I still doubt that we
> can eliminate it completely without totally breaking the underlying
> principals of Freenet's operation.
If you have the resources to port scan blocks of IPs (which I think is a
reasonable power to give a theoretical attacker) then you can't make node
fishing less effective. I stand by my position that the option to reject
unknown connections is the only way to combat this attack and that it will
not break Freenet.
_______________________________________________
Freenet-dev mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/mailman/listinfo/freenet-dev
