One hop proxy [Re: Can nickname be duplicate?]

Pei Hanru peihanru at gmail.com
Thu Jan 31 08:17:57 UTC 2008


> Before tor is available, people often used proxy located abroad their own
> country to visit banned sites, but this was a unstable way, such proxy
> server was hard to find, once one of these proxy become well-known, it would
> soon be listed on the banned IP list of censor system. Furthermore, it
> cannot avoid "key-words" filter if data streamed from client to proxy (or
> vice-versa) is not encrypted.
> 
> The solution is a proxy without fixed IP and can also encrypt data, Tor is a
> good tool! But we just need such a proxy, thus so many relays is not needed,
> and such relays will slow down the speed of communication.

There're already some one-hop-proxy tools for this specific purpose (I'm
not going to mention them publicly and assuming you're familiar with
them, if not, please drop me a private mail), and generally they're
faster than Tor.

> So, my idea is to find a way to get rid of relays, what I need is just exit
> nodes abroad my country and other totalitarian governed regions, I've found
> that generally one circuit contains three nodes when tor is used to browse
> website, that is to say my data is encrypted for three times. In fact, to
> me, one exit node with a high bandwidth abroad is enough.

There will be a lot of problems with one-hop circuit. For example, if
"bad" people know that there are one-hop circuits in the Tor network,
they will be more interested in attacking Tor, setting up more malicious
nodes, etc., the consequence is a mess.

I agree with you that Tor is slow, especially in China, we need to
improve this, but alas, it is easier said than done.

> Of course what I have said above does not necessarily fit every country. In
> my country, as long as you do not spread out those banned information, the
> police would not bother you just for the reason that you browse them
> personally.

I'm also from China, haven't investigated such "browsing unsuitable
material but not spreading" issue you described. How do you draw the
above conclusion? Is it written somewhere?

Hanru



More information about the tor-talk mailing list