<DIV>Well, count me in as your first "trendy" subscriber... I'd don't have a clue about writing code and I'm really too damned boring to need privacy, but I'm trying to figure out how to participate as a server just to be part of the "movement"...<BR><BR><B><I>Matt Thorne <mlthorne@gmail.com></I></B> wrote:
<BLOCKQUOTE class=replbq style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #1010ff 2px solid">I can see where requiring bandwidth for bandwidth usage would fail...<BR>But I just didn't want to create something that was this rediculusly<BR>easy to use w/out some more redundancy on the network. Don't get me<BR>wrong, tor is still easy to use, but this is over the top above and<BR>beyond easy. people who didn't really need anonymity would still use<BR>it. it might even become... "trendy" (gasp)<BR><BR>-=Matt=-<BR><BR>On 11/19/05, Nick Mathewson <NICKM@FREEHAVEN.NET>wrote:<BR>> On Sat, Nov 19, 2005 at 01:40:56PM -0600, Arrakis Tor wrote:<BR>> [reformatted to fix top-post.]<BR>> > On 11/19/05, Nick Mathewson <NICKM@FREEHAVEN.NET>wrote:<BR>> > > On Sat, Nov 19, 2005 at 12:45:48PM -0600, Arrakis Tor wrote:<BR>> [...]<BR>> > > > How do routers do it with DNS tables, that we can't with Tor servers?<BR>> > ><BR>> > > DNS maps names
to values, and doesn't worry about vulenrabilities<BR>> > > resulting from adversary knowing which clients have learned which<BR>> > > values. That's not our problem. Our problem is finding a way for<BR>> > > clients to learn about servers and build paths through those servers<BR>> > > so that if you (an adversary) see a client, and you control a<BR>> > > directory cache, and you control some servers, and you see part of the<BR>> > > client's path, you can't deduce with a better-than-chance probability<BR>> > > whether the path was generated by the client.<BR>> ><BR>> > No no. I mean how routers propagate routes and know paths. Inherently,<BR>> > a router does not know the entire path of the internet, however it<BR>> > does know who is around, and how to get where it wants data to go.<BR>><BR>> Ah. I was thrown off by the fact that you said "DNS", not "BGP" or<BR>> something. DNS has
nothing to do with how routers learn paths on the<BR>> Internet, so I didn't know you were talking about how routers learn<BR>> paths on the Internet.<BR>><BR>> Once again, BGP solves a very different problem: how to make sure that<BR>> each router knows the best way to send an IP packet closer to where it<BR>> is supposed to go. It doesn't concern itself with the anonymity<BR>> questions I mention above in the slightest.<BR>><BR>> Trivial example: In Internet routing, you typically trust the first<BR>> router you use to route your packets. But in Tor, if you trust the<BR>> first router to pick your path, or give you a list of routers, you are<BR>> completely vulnerable to a compromised first router. There might be<BR>> ways around this problem (and the other problems) but they need design<BR>> and analysis.<BR>><BR>> yrs,<BR>> --<BR>> Nick Mathewson<BR>><BR>><BR>><BR></BLOCKQUOTE></DIV>