On Wed, 2004-08-04 at 16:35, Toad wrote:
> On Wed, Aug 04, 2004 at 04:31:11PM -0400, Edward J. Huff wrote:
> > On Wed, 2004-08-04 at 16:00, [EMAIL PROTECTED] wrote:
> > > I'm not sure where your 'village' is but here it works much 
> > > the same way actually.  But the problem is that there is no 
> > > machine that can just tell us what your intent was.  So what 
> > > your intent was has to be inferred from your actions and your 
> > > knowledge.  The fact is that everyone knows there lots of 
> > > illegal stuff floating around freenet, and one can simply 
> > > not avoid responsibility for a crime by deliberately ignoring 
> > > what is obvious.
> > 
> > > So even though you didn't want to transmit kiddy porn you 
> > > made the choice to run a freenet node fully aware that it 
> > > could and would result in KP being distributed.  That right 
> > > there is enough to establish intent.
> > 
> > Ok, suppose most users of freenet decide to unite against
> > kiddie porn by using TFE, YOYO, etc., to learn as many KP keys as
> > possible, and delete these keys from their datastores and patch freenet
> > so it won't carry them.    Now even so, some KP will be distributed, but
> > only so long as the keys are unknown to the general population of
> > freenet users.  Now what do you say about intent?
> What happens when users start deleting less obviously problematic files
> such as warez and mp3z? What happens if they disagree over what should
> be deleted? And as far as child porn goes, don't you think a lot of it
> will be "underground" i.e. not readily available from TFE? There was an
> IIP board dedicated to such things... Anyway, if we start self
> censoring, we have two problems:
> 1. Everyone will have a different idea of what should be censored.

In that case, freenet will route around the most restrictive nodes.
Censorship will only be effective if a clear majority of nodes elect to
censor the content.

> 2. Anyone who censors child porn but not warez, or warez but not decss,
> or decss but not $cientology copyrighted papers, can be compelled to
> censor the rest, since it is also technically illegal. 

That is not a problem with my suggestion.  It is a problem with the
fundamental design of freenet.  A system which avoids this problem would
have to make it impossible to tell at all (not just impossible to be
100% certain) who is requesting content and who is supplying content. 
It would have to be impossible to tell what content is passing through
each node.  ("Impossible" means without compromising a substantial
fraction of all nodes).  

Freenet does not achieve this, except when the crypto key (the part
after the comma) is not published.  Once the crypto key is published, it
is no longer impossible to tell what is passing through the node.

> If we are to "cooperate" in the sense you suggest, we cannot simply 
> block child porn.  We would have to block *anything that is illegal
> in the node op's jurisdiction* !

That is up to each node operator.  Failure to block some content -- like
mp3's -- is a lot less serious than failure to block other content --
like kp.  The node operator might decide to take the risk in the name of
civil disobedience for some content but not other.

This decision _is_ forced upon the node operator by the design of
freenet.  A different design might avoid the problem by making it
actually impossible to do selective censorship.

Attachment: signature.asc
Description: This is a digitally signed message part

Support mailing list
Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support

Reply via email to