At 11.13 05/02/02 -0500, you wrote: > > > On Tue, Feb 05, 2002 at 12:21:18AM -0500, Gianni Johansson wrote: > > > I added a ClientDelete command to the node's FCP implementation to delete ... > > > > But why would we make it easier for people to censor their own > > datastores? > > >In a well functioning network ClientDelete should be almost useless as a >censorship tool because of distributed caching of popular data. > >It is for people (organizations? ;-) ) like CofE who are inserting from >transient nodes. > >Deleting keys from the local store and re-requesting them allows freesite >publishers to see whether their content actually made it out into the network. > >All I am doing is providing a scalpel to substitute for the sledgehammer >approach that most successful transient inserters are already using -- i.e. >nuking the data store.
AFAIK the Freenet architecture, this tool is an useful debugging and administration tool. But AFAIK the security model of Freenet (difficult matter due to lack of docs) this is just an hole in the "deep security" model of Freenet as a whole; I cannot say why, because is just another weapon for an attacker, but I can say that not giving one more weapon is better. Justifying this choice for usefulness is possible (but the usefulness is not big, IMHO), but it is impossible give a proof of impossibility to use this function in an attack. No way to delete information (other that nuking the node), this is a thick wall against an attack. JM2C. Marco * Marco A. Calamari marco at freenetproject.org * il Progetto Freenet - segui il coniglio bianco the Freenet Project - follow the white rabbit _______________________________________________ Devl mailing list Devl at freenetproject.org http://lists.freenetproject.org/mailman/listinfo/devl
