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

[George Kadianakis]

# Goals of February 2014

- I haven't worked on this as much I would have liked, but it's one of
  the most important news of the previous month for me:

  David's PTTBB patches _and_ Kevin's FTE patches got merged in
  tor-browser-bundle.git! This means that we now have gitian builds
  for TBB that supports pluggable transports. Good news!

- Also, ScrambleSuit was merged in obfsproxy! We did an announcement
  in tor-relays [0] and a few bridge operators have setup scramblesuit
  already.

  FWIW, a few days ago, Yawning found a few bugs (#11100, #11092) in
  the scramblesuit codebase that we will fix soon and roll a new
  release of obfsproxy.

- As part of the new obfsproxy release and announcement we got several
  mails of confused people asking reasonable questions. We hopefully
  addressed most of them and made tickets for the issues that we
  should fix (#11043).

- Somehow I got involved with the improved guard parameters project
  that has been floating around for a while [1]. We had several
  discussions about it in the dev meeting and we made a few plans on
  how to move this forward [2].

  On this note, I started evaluating the effects that such a change
  would have to the Tor network (both from the anonymity and
  performance PoV) [3]. So far that thread has generated a few useful
  graphs that we need to analyze and tie together to answer a few
  questions:
  * How much will that change influence the performance of the
  network?
  * How much will that change influence the diversity of the guard
    node selection procedure?
  * If we do this change, what should be the new parameters so that we
    maximize the performance of the network without harming its
    anonymity?

  Nick is planning on writing a little-t-tor proposal for these design
  changes. I plan to help him with that.

- mikeperry informed us of a bug in Tor that made it hard for
  tor-launcher to understand whether Tor (configured with bridges)
  failed to bootstrap because of network issues or because it's still
  bootstrapping (#11069).

  We evaluated the source of the problem (i hope) and provided some
  patches to fix it (i hope). Said patches will be merged within the
  0.2.5.x merge window (i hope), so that's good news for TBB.

# Goals for March 2014

- Fix #11100 and #11092 and do a new obfsproxy release.

- Include scramblesuit and FTE in TBB and get them deployed to real
  users.

- Look closer at meek: David's shiny new transport [4]. See what needs
  to happen before deployment. Start planning for deployment.

- Work on the guard stuff. Finish the analysis, decide something
  useful about the parameters we should use, help Nick with the
  proposal and start working on it.

- Reopen rend-spec-ng.txt and specify how HSes are supposed to pick
  IPs and how long they should keep them for. Also specify how clients
  are supposed to pick RPs.

- Look more at obfsclient. Look at the UniformDH spec issues that
  Yawning has pointed out (#10782).

- Hope for some good GSoC students :)

[0]: https://lists.torproject.org/pipermail/tor-relays/2014-February/003886.html

[1]: https://blog.torproject.org/blog/improving-tors-anonymity-changing-guard-parameters
[2]: https://lists.torproject.org/pipermail/tor-dev/2014-February/006288.html

[3]: https://lists.torproject.org/pipermail/tor-dev/2014-February/006289.html

[4]: https://trac.torproject.org/projects/tor/wiki/doc/meek