[tor-bugs] #11629 [FTE]: FTE breaks with latest obfsproxy (because of move to SOCKS5)

[Tor Bug Tracker & Wiki]

#11629: FTE breaks with latest obfsproxy (because of move to SOCKS5)
------------------------+--------------------------
     Reporter:  asn     |      Owner:  kpdyer
         Type:  defect  |     Status:  needs_review
     Priority:  normal  |  Milestone:
    Component:  FTE     |    Version:
   Resolution:          |   Keywords:
Actual Points:          |  Parent ID:
       Points:          |
------------------------+--------------------------

Comment (by kpdyer):

 Replying to [comment:3 yawning]:
 > Replying to [comment:1 kpdyer]:
 > > I assume that there is a proper way to use pyptlib and my
 implementation is just a cheap hack?
 >
 > There's two correct ways to use it.
 >
 > One would be to write your own networking code to provide the
 infrastructure being pulled in from obfsproxy (The equivalent of the
 `obfsproxy.transports.base.BaseTransport` code).

 Was hoping to avoid that!

 > The other would be to do what ScrambleSuit did and use obfsproxy for
 everything (See #11354).  This will probably be easier from a development
 standpoint, but there is a licensing conflict since all the FTE code is
 GPLed, and I'm not sure how asn feels about adding a dependency to libfte
 that includes native code to obfsproxy.

 For now, it wouldn't make sense to include FTE in obfsproxy. Predominantly
 because fteproxy has some C++ that would complicate the obfsproxy
 deployment process, which is currently pure-python.

 -Kevin

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/11629#comment:4>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online