[tor-project] Network team meeting, 7 August 2017

[Nick Mathewson]

Hello, friends!

We had another network team meeting today.  The log is at
http://meetbot.debian.net/tor-dev/2017/tor-dev.2017-08-07-17.00.html

Notes from the pad are below. A lot of people are out for various time
periods during August.


Network team meeting, 7 August 2017

Notes from last week:
    https://lists.torproject.org/pipermail/tor-project/2017-July/001337.html

Reminders:
    - 0.3.1.x was scheduled to be stable last week; nickm postponed it
one month to 5 September.
    - 15 September is the 0.3.2.x freeze.

Nick:
    * Last week
      - Finished and circulated draft of privcount-in-tor proposal to
privcount authors; updated code to match.
      - Finished reviewing #20657 (prop224, server-side); reviewed
first round of fixes.
      - Reviewed and merged lots of branches, fixed lots of regressions.
      - Released 0.3.1.5-alpha, 0.3.0.10
      - Announced the end of 0.2.4, 0.2.6, and 0.2.7.
      - Wrote a bunch of little patches
    * This week
      - Send privcount draft to tor-dev; open tickets; etc.
      - Review and comment on tor-client-progress draft.
      - Review #20657 2nd round of fixes, and maybe 3rd?
      - Get ready for being out of town next week. Anything I should
do/know this week?

 - More hacking, probably.

 - Should I do 0.3.1.6-rc this week?

   (No. Once back from vacation.)


 komlo:
     - This week:
         - first round of rust protover code review by isis
         - write fuzzing tests for protover ffi

- Next week:

- Schedule for rust hackfest


asn
  Last week:
   - Fixing up #20657 branch from nick's review.
   - Testing #20657 and finding more bugs.
   - Wrote tons of #20657 unittests.
   - Did some testing/reviewing on the client-side branch
  This week:
   - Get #20657 merged hopefully.
   - Start reviewing the client-side.

pastly:
    Updating KIST tickets as we speak.
    David: KIST will be needs_review by the end of the day



catalyst:
    Last week (2017-W31):
        - helped review #23061 (crypto_rand_double()); floating point is HARD
        - looked at #22636 log output on failure -- seems OK actually
due to the way automake's test rules extract failing/skipped detailed
logs into test-suite.log.
        - dealing with myriad small annoyances of a new house
    This week (2017-W32):
        - copy brade's comments about Tor Launcher behavior from
#22232 to somewhere more useful like a wiki page
        - think some more about #23061
        - think some more about how Core Tor can help Tor Launcher
with bootstrap troubleshooting

isis:
    last week:
        - began writing up a specification for Ristretto (Decaf point
encoding for curve25519), this way we can agree that we're all doing
the same encoding, and then I can use it in Hyphae (#22776)
        - revising patches for the new cmdline key expiration feature (#17639)
        - some torspec patches
          https://gitweb.torproject.org/torspec.git/commit/?id=8ebdf2e99f93628055c7d7a89327191cf4d161dc
          https://gitweb.torproject.org/torspec.git/commit/?id=28a8208232cac218b9bbbd95521ad7bd3c266b37
          https://gitweb.torproject.org/torspec.git/commit/?id=db17344f15a9e02d041e9d51ff325133cc07007e
        - laptop had hardware failures and needed to be repaired and
then reinstalled :(
        - also starting building a desktop/build/work machine as well,
but that's not finished yet
    this week:
        - reviewing Chelsea's Rust protover patch
        - more work on the moat backend server (#22871) and farfetchd (#15967)
        - getting ready for my talk at rustconf next week