[tor-reports] George's status report: April 2014

[George Kadianakis]

# Activities of April 2014

- The obfsproxy/pyptlib garden was lively again this month [0].

  Fresh fruits and vegetables emerged with Yawning's help. Specifically,
  we merged Yawning's code for outgoing proxy support just in time to
  realize #11658. We also released obfsproxy-0.2.9 so that we can
  include it in the upcoming 3.6 bundles that will solve #11658 by
  including the little-t-tor fix to #8402.

  I also prepared a branch that merges the latest fixes to
  scramblesuit that haven't been applied to obfsproxy yet (#11271).
  Yawning reviewed and I'm also waiting for phw's review, then I will
  make the necessary changes and merge them to obfsproxy to be
  included in obfsproxy-0.2.10 .

- Solved a few PT-related issues in little-t-tor.

  During the work on BridgeT with Arturo, we found an issue with my
  fix of #11156 that would make bootstrap hang for 60 seconds. We have
  since fixed the issue.

  My patch to #11069 also had a bug (worst coder ever), where it would
  log "%s" instead of logging an error message (it was pretty close to
  memory corruption too...). We fixed that in #11654.

  wfn started working on #9957, which will make PTs easier to debug
  (it will print in the Tor log any ImportError or other fatal error
  messages that PTs spit in stderr). I did an initial review of his
  code.

- Continued work on the single guard proposal.

  Nick Hopper helped me [1] understand his idea on fixing the problem of
  the underutilization of new guards [2].

  I published some notes on an implementation plan of the proposal:
  https://people.torproject.org/~asn/guards2/single_guard_impl_plan.txt

  I hear there might even be an upcoming tech report on the matter.

- Accepted Juha as a GSoC student on the topic of
  "GSoC: Ahmia.fi - Search Engine for Hidden Services".

  The idea is that Juha will work on improving the ahmia.fi hidden
  service search engine. You can read a potential time schedule of his
  summer at https://ahmia.fi/gsoc/ .

  Always fun to have Google fund sister search engines :)

- Got hit by the heartbleed [3] craze big time.

  Spent the first two days of disclosure helping enumerate vulnerable authorities/relays.
  I patched my own bridges/relays.
  Spoke with people who run busy obfsbridges and asked them to patch their systems.

# Activities for May 2014

- This month will be quite intense towards my non-Tor life. Expect to
  see less activity, but by the end of the month it will return back
  to normal.

- See what's up with #11658 and whether the upcoming TBBs will fix the
  problem with outgoing proxies and PTs.

- With regards to the single guard proposal, Nick suggested that we
  should do some refactoring of the entry guards code while
  implementing the proposal because it's a bit rusty. I plan to look
  at the code and see what can be refactored this month. The idea is
  to start touching code by the end of May.

- The previous obfsproxy release (0.2.8) broke FTE which was importing
  some internal networking code from the guts of obfsproxy.

  It slipped my mind that replacing obfsproxy's SOCKS4 with SOCKS5
  (#8956) would cause an issue, and that gave birth to #11629. It
  would be nice if we found a nice solution for this issue.

(This status report was a bit lengthy. If that causes you to ignore
it, please tell me and I will make it briefer. If you ignore it
anyway, you know the secret.)

[0]: https://upload.wikimedia.org/wikipedia/commons/c/c9/Abel_Grimmer_001.jpg

[1]: https://lists.torproject.org/pipermail/tor-dev/2014-April/006654.html

[2]: phase three of https://blog.torproject.org/blog/lifecycle-of-a-new-relay

[3]: just joking