[tor-bugs] #4248 [Analysis]: Get better at comparing SSL speed in Iran to Tor speed

[Tor Bug Tracker & Wiki]

#4248: Get better at comparing SSL speed in Iran to Tor speed
----------------------+-----------------------------------------------------
 Reporter:  arma      |          Owner:     
     Type:  task      |         Status:  new
 Priority:  normal    |      Milestone:     
Component:  Analysis  |        Version:     
 Keywords:            |         Parent:     
   Points:            |   Actualpoints:     
----------------------+-----------------------------------------------------

Comment(by arma):

 Replying to [comment:2 runa]:
 > Define "guru"? Sebastian, Robert and I have all set up/configured
 stunnel and we might be able to answer some of your questions.

 What we need are specific instructions that can be run on the client side
 and the bridge side.

 On the client side, we want to set up an stunnel that accepts incoming
 connections to localhost and forwards them, encrypted, to the remote side.
 Then the Tor client should configure that stunnel local socket as her
 bridge.

 On the remote side, we want to accept incoming ssl connections, decrypt
 them, and then hand them to the local Tor relay. For example, fork a "nc
 localhost 9001" for each incoming connection after the handshake. That way
 the connection from the Tor client goes to the Tor bridge, but it
 transparently has another layer of encryption added and then removed in
 transit.

 We'll want to try this on Tor bridges, not relays, due to bug #1776.

 As a bonus, if this works I bet it will make bridges usable in China again
 (for now).

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/4248#comment:4>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online