Re: Crazy idea: How trust in darknets enables secure democratic censorship

Matthew Toseland wrote:
> On Wed, Jul 13, 2005 at 10:33:59AM +0200, Matthew Exon wrote:

>>Now, eventually everyone's going to behave the same way.  Everyone will 
>>figure out that there's just no point posting star trek episodes to the 
>>nasty darknet, since they'll always be more easily available on the nice 
>>darknet.  Result: the nasty darknet will have nothing but dead chicken 
>>porn.  There's no reason to post anything else there, and no-one will 
>>ever download anything else from it.
> 
> 
> It is quite possible to manually migrate content between darknets (without
> being the content author, and without disrupting the linkage).

Sure.  But why would I do that?  The only reason would be if I believed 
that lots of people who aren't on the nice darknet would want to see it. 
  You're right that some people wouldn't participate in both networks 
because of bandwidth constraints, but I can't help feeling that this 
would be a fairly small number.  There's no hard lower limit on how much 
bandwidth you need, especially with the new version; so no matter how 
little you have you can always share it.

>>My question is: how secure is the nasty darknet now?  It seems that much 
>>of the security or plausible deniability of freenet stems from having a 
>>good quantity of legitimate content alongside the punishable content. 
>>But now the nasty darknet is 100% dead chicken porn.  If the authorities 
>>can demonstrate that I participated in that particular darknet at all, 
>>I'm screwed.
> 
> 
> Hmmm. Plausible deniability, yes. Security, no.

Fair enough.  Am I right that plausible deniability for the nasty 
darknet is pretty much blown?  Is this a big problem?

>>(Sure, I know that you're assuming freenet is illegal; but there's 
>>illegal and then there's illegal.  An Al Qaeda darknet will have a lot 
>>more resources thrown at it, and a lot more rules bent, in order to 
>>uncover all the participants than, say, an MP3 sharing network.)
> 
> 
> Al Qaeda could use the main darknet, and make sure nobody found their
> content. That is if they used freenet at all. This option is open to any
> sufficiently closed group, of course, but they are vulnerable to a
> disgruntled member (or a mistake).

OK, Al Qaeda just decided to do precisely that.  So now I, as a nice 
darknet participant, have to live with the knowledge that I'm helping Al 
Qaeda plan their terrorist attacks.  We're back where we started aren't we?

In other words, for the censorship system to make a difference, there 
has to be an incentive not to *hide* content on darknets that don't want 
it.  It's not enough to just punish the posting of open content.

>>So it seems to me that any dead chicken porn fetishist reading this 
>>discussion should be totally opposed, because it would leave him more 
>>exposed.  And the whole point of Freenet is to be there for persecuted 
>>minorities, right?
> 
> 
> A small darknet may actually be safer in terms of its likelihood of
> exposure. If the state randomly searches people's computers, obviously
> you're in trouble no matter what.

Yes and no.  The scenario I was thinking of is the agent provocateur. 
Guy I meet down the pub offers to bring me into a darknet, but of course 
he's actually a cop.  If it's a general darknet without any particular 
focus, no-one can know for sure that I set out to find dead chicken 
porn.  But if I specifically needed a dead chicken porn darknet, then 
this would have come up somehow when I was talking to the cop, so it's 
rather more serious.  It still seems like trying to find dead chicken 
porn is going to be far riskier with the censorship system in place than 
without it.

> Depends what you mean by discriminating against, I suppose. Minorities
> who the majority vehemently opposes would have to form their own
> darknets. When they try to join with the main darknets, they'd be
> expelled en bloc.

What I mean by discriminating against is, where does this lie on a 
continuum from "has no effect on" to "making it more difficult" to 
"making it impossible"?  Ideally, it wouldn't be any harder to find 
minority material than to find majority material.

> I doubt that even a censorable freenet would ever be
> so mainstream that what is mistakenly illegal in real life is always
> voted down - but if it is we can always build an alternative community.

For purposes of this discussion, I think it's worth each of us assuming 
that we belong to the minority, not the majority.  The majority really 
isn't the interesting bit of Freenet.

To be specific: China.  China has a pro-democracy movement.  It might be 
illegal, but it's very popular, and very widespread, especially in the 
technology-rich East.  Any Freenet community that gets built up in China 
is likely to be based around pro-democracy Han Chinese students and all 
their computers.  Due to the great firewall of China, a strongly 
connected Chinese darknet would form, weakly connected to the rest of 
the world. (The firewall can't stop Freenet, but it sure as hell slows 
everything to a crawl.)

China also has a Tibetan independence movement.  Not nearly as many 
people as the massive Han-based pro-democracy movement, and very few 
have access to computers.  On the Chinese darknet, Tibetan independence 
advocates are the tiny despised minority.  And anyone who thinks that 
Chinese students are libertarian should try talking to them.  Chinese 
pro-democracy advocates are, if anything, *more* nationalist, racist and 
paternalist than the government.  I really don't believe they would 
think twice about turning in "separatists" to the government, just as I 
wouldn't have too much trouble turning in Al Qaeda members.

The Tibetan activists are trapped behind the Chinese firewall with all 
the Han Chinese.  They need to be allowed on to the Chinese darknet so 
they can communicate with their friends in India and the West.

So the question is, how do we ensure that the Tibetan activists can 
participate on Freenet without unnecessary fear of reprisal?  Does the 
censorship system put them at risk?  I still don't know if it does or 
not, but I certainly don't have a lot of faith in the view of the 
majority, even the Freenet majority.