[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Mon, Mar 24, 2014 at 02:01:44PM +0000, Philip Hudson wrote: > The insight comes from holding not just links to data but the actual > data itself, right? Meaning the data can be quickly searched, indexed, > analyzed etc on every peer, quickly as in without that peer having to > pull it in over the Net. That seems to me be both an opportunity and a > risk, and rather more of an opportunity at that (in most cases). > > So I guess I can see how an adversary could gain this insight, but I'm > not sure a) it matters or b) it's what the proposal is meant to > prevent. I'm not an expert on censorship, but typically you want to 1) access the content that has been blocked and 2) prevent the censor from finding out you've accessed this content. The proposal is a neat way of achieving 1). But not only does it nothing to achieve 2), it could make the situation worse. In almost all cases, this matters: if you're accessing an opposition website while living under a totalitarian regime, you don't just want to evade the blocking of the website, you also want to make sure the government doesn't find out you've accessed the website. And by running some peers, the government can easily find out who has been accessing said website. It doesn't even have to happen at a large scale: it suffices for people to know this can happen for them not to access the data anymore. Hence the cilling effect. Martijn. -- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/listfaq