What I'd like to see is a way to enforce restrictions on the destinations you're willing to route as an exit node. I'm all for giving people living under restrictive regimes access to things like Wikipedia, but I don't particularly care to enable anonymous access to child pornography.
Either you have freedom of information, or you don't. All you're saying is that you want to write the rules yourself.
edit: What I mean, is it's a slippery slope. If you do the selection by yourself, you'll let out a lot of good stuff. If you trust somebody else to do the selection, that someone will be in time important enough to become either corrupt or a target, or both. Just imagine google excluding sites from its index on some subjective criteria.
You can't solve this kind of problem by limiting access to information. It's simply a bad strategy from the start, and like any design mistake it creates problems farther and farther away, when the issue is only one: a bunch of bits have been made illegal.
I strongly disagree with this idea that I've seen a few times, that somehow the law is irrationally and ridiculously making bits illegal whether in copyright or in pornography. Child pornography is illegal whether its printed on paper or stored as bits. That somehow it shouldn't be a crime just because of the fact that bits can be used to represent anything is a really bad argument.
Well, obviously a blacklist wouldn't work, so are you suggesting a whitelist? Seems like that would seriously reduce the usefulness. How would you know what foreign-language dissident news sites to include?
What about the ability to spread out a particular request to many exit nodes? This would probably require the establishment of new formats and protocols. How about image formats with the same sort of information as Bittorrent? (Can someone be prosecuted for downloading a fraction of something?) As mentioned in the Rivest article posted recently, there are also "packaging transforms" that allow one to divide information up into n parts, such that all n parts must be provided to retrieve it.
I am surprised that such protocols haven't been devised yet. Yes, this would enable those engaged in child pornography. But those so engaged will just concoct yet another mish-mash of obscurity in any case. (Like using encrypted Windows Remote Desktop connections to virtual servers with encrypted archives on them.)
Just because Bittorrent is used by so-called "pirates," doesn't invalidate it or its use to distribute ISOs of Linux distributions. Likewise, such protocols would patch TOR's weakness at its exit nodes.
Yeah, it's pretty disheartening that a guy who really was trying to do something good for folks who have restricted legal access to the Internet, is hammered by the cops because of some douche-rocket that needed his pedophilia fix.
In his case, maybe white-listing would have been a good idea. I don't know what a Tor directory looks like, but perhaps there could be a list of Tor nodes that give access to specific sets of sites.
I agree with you. I would be extremely uncomfortable knowing that I might be enabling access to child porn.
But you might have problems legally if you take any responsibility for the kind of traffic that goes through your node. If you try to block anything, that might be implying legally that you have some responsibility for the traffic that goes through the node.
So a whitelist would be best, and you better be sure that every single thing on the sites on your whitelist is legal in your home country. For example, Australia has censored a wikileaks page. So you wouldn't want to allow access to that wikileaks page if you were in Australia.
Right now, you can use Google to find pirated material. You can even limit the file type to torrents. Google's main defense is that it has legal uses, and it is not responsible for policing files on other sites.
