[tor-project] Network team meeting notes, 24 July 2017

[Nick Mathewson]

Hi!  Below's the pad from this week's meeting.

The meeting transcript is available at
  http://meetbot.debian.net/tor-dev/2017/tor-dev.2017-07-24-17.00.html


=====

Network team pad, for 24 July Meeting (or 25 July, for those in UTC+5 or later)


Notes from last week's meeting:
    * https://lists.torproject.org/pipermail/tor-project/2017-July/001305.html
      (Did we do what we had planned?)

Announcements and reminders:
    * Various network team hckfests are on 9th to 16th October. Nickm
is confused about when.

    - we should make a pad for the schedule isis wrote, with updates (komlo)

    * I've added the August calendar for the Team Rotation -dgoulet
    * Not much progress on review-group-21 :/

Things we should talk about:
    *

Tasks for after the meeting:
    *

teor (not online):
    Last week:
        * Took a few days leave, because ugh, winter, again
        * Wrote some more tor circuit code for endosome, found some
tor-spec ambiguities
        * Did a code review on some of Rob's PrivCount code
        * Booked Montreal travel and arranged hackfests and meetings
        * Worked out when the grant money for my current job will run out
    This week:
        * Implement one final PrivCount feature before deployment
        * Test PrivCount before deployment
        * Deploy and Run PrivCount Single Onion Service counts
        * Maybe get cell crypto working in endosome

taylor:
    (I believe this is the week that Taylor is moving. Let's all wish
them an easy relocation! -nickm)

asn
  Last week:
   - Wrote patch for #22735. Found and fixed some bad bugs on #20657 while
     writing this patch.
   - Wrote code and tests for #22940.
   - Continued review and testing of #20657. Found some more issues,
     particularly regarding the overlap mode and descriptor rotation logic.
   - Reviewed #22895 and #22979.
   - Did some hackerone bug triaging since we publicly launched our bug bounty
     campaign.
  This week:
   - Mod HS circuitmap to be used by client side prop224 for RP circs.
   - Continue reviewing and testing of service-side #20657. Find and fix more
     bugs, and hopefully move it towards upstream merge.
   - Start reviewing David's client-side code.
   - We are approaching prop224 merge time, so do any other prop224 action
     items that must happen.

nickm:
    Last week:
        - PETS. Had good convesations about
          . A couple of interesting attack papers
          . anti-fingerprinting research
          . traffic padding
          . The research safety board
          . I2P
          . getting large patches/add-ons deployed
    This week:
          . Review, merge, bugfix. :(
          . Document, specify :(

        . Work on guard stuff, measurement, trace stuff. (dgoulet:
lttng advice? Also I should coordinate with teor)

          . Enter sadly delayed sponsor8 items from meeting with ahf
onto bugtracker so I can't forget


ahf (might be missing the start of the meeting):
    Last week:
        Sponsor 8:
        - Up and running with Android development environment for
Orbot + Core Tor.
        - Up and running with Android power measurement via their
dodgy Python scripts (on a nexus 5x).
        - Met with a friend in Copenhagen to get a fast walkthrough of
the Android environment.
        Misc:
        - Did a talk on Tor at TheCamp.dk.

    This week:
        - Document Sponsor 8 development environment with Android on trac wiki.
        - Figure out how we can collect measurement on Android about
          battery/CPU usage where we have "fast" test-to-result times.
        - Create tickets for sponsor8.
        - Look into our wake-ups on an Android device.
        - Figure out how much we can measure CPU wake-ups/network via the
          Android simulator.

    Things I could need help with:
        - Find Nick + Catalyst's pad/log about control port enhancements (URL?)
        - Need to figure out what functions in core tor that does network
          activity on timers.

dgoulet:
  Last week:
    - Massive amount of prop224. Addressing asn's review of #20657
    - Mostly coding #17242.
    - Prop224 is now closer to reality where we had our first client<->service
      connection yesterday ;). The whole chain has been glued together and is
      working!
  This week:
    - Finalize #17242 (client code) so we can test the whole stack more
      thoroughly.
    - Need to make a branch that contains it all for arma so he can showcase
      it at our *last* final sponsor R meeting and then two days later to
      Defcon.
    - Basically, expect me to only focus on prop224 so we can reach our
      deadline goal and start merging more and more upstream.

komlo (not online):
    - I am starting my new job this week (!!), so will pick up more
tor work next week

pastly (also attending meeting in #tor-project for people who provide support):
    - Measuring latency in Tor seems to be going well. Reimplemented
Ting all last week.
    - Now that Rob is back from PETS, will discuss how to update KIST tickets.


Mike:
  Last week:
    - PETS
    - Met with Matt Wright, Marc Juarez,  Moshen Imani, Nick, Roger
about prop254
      - Discussed potential issues with flow control and other planned
histogram changes
    - Worked with Matt on outline of stuff for Sponsor2 report. (He
will be submitting it.)
  This week:
    - Catch up on mail and other things since PETS
    - Flesh out details for Sponsor2 report; send them to Matt
    - Work on torrc options for Prop247 prototype (#13837)
    - Help with Tor Browser interview process

Isabela:
    Last week:
        - PETS
        - Finished the modularization proposal - submitting as we meet
(today, cuz we were waiting to hear back from Laura on a question
related to the submission form)
    This week:
        - NSF reports! help organize that for arma (Mike already
created a plan for sponsor2 one while at PETS)
        - We got sponsor8 contract!! I will get  you more info as we
figure out if we want our start date to be August 1st or later
        - July is ending and I wonder if we are closing the
deliverable for sponsor 4
        - I sent an email to try to create a backup person plan for
vacation season - sent to tor-internal.

isis:
    last week:
        - sending out emails for internship and getting things set up for them
        - discussions about differences in signatures #16562
        - upgraded bridgedb server to Debian 9 and recalibrated the CI #22998
        - did some thinking about potential security implications of
not having random payloads in drop cells #22948
        - revised tor.git travis CI configs #22636
        - looked into dependency libraries for intern project
    this week:
        - revise captcha server for moat according to review/specs #15967
        - get the bridge bandwidth scanner project rolling