[tor-project] Weekly network team meeting notes (20 Nov 2017)

Nick Mathewson nickm at torproject.org
Mon Nov 20 18:06:40 UTC 2017


Hi, all!

You can find the logs of our latest weekly meeting here:
  http://meetbot.debian.net/tor-meeting/2017/tor-meeting.2017-11-20-16.59.html


Below are our notes:

==============
Network team meeting pad, 20 November 2017

Welcome to our meeting!  Every monday at 1700 UTC on #tor-meeting on
OFTC.  (This channel is logged while meetings are in progress.)

Want to participate?  Awesome!  Here's what to do:
    1. If you have updates, enter them below, under your name.
    2. If you see anything you want to talk about in your updates, put
them in boldface!
    3. Show up to the IRC meeting and say hi!

Note the meeting location: #tor-meeting on OFTC!
 (See https://lists.torproject.org/pipermail/tor-project/2017-September/001459.html
for background.)

Meeting notes from last wek:
    * https://lists.torproject.org/pipermail/tor-project/2017-November/001568.html

Announcements:
    - Remember, it's a US holiday on Thursday. Some US folks will be
on limited availability Wednesday afternoon through Monday.
    - Sponsor M is no longer billable: please let me or isabela know
if you need to do more work on M stuff, and we can work something out.
    - On the roadmap spreadsheet: Please start taking december/january
tasks.  (If somebody else has already taken something you want, please
talk to them and/or add yourself too.)

Discussion Topics:
    - 0.3.2 planning
    - Time allocations

dgoulet:
  - Last week:
    * Various 032 tickets work (review and patches).
    * Finalized the hs v3 control port support (#20699). Missing some unit
      tests before ready for review.
    * Worked on TROVE-2017-012
    * A bit of thinking on channel/scheduler problem especially how to better
      handle DESTROY cells. I've started writing a proposal (not finished).
      (https://lists.torproject.org/pipermail/tor-dev/2017-October/012536.html)
  - This week:
    * Making #20699 in needs_review.
    * As usual, ticket work on 032 milestone.
    * Tackle #23709 which is the removal of the in/out queues from channels.
      This will need much more unit tests than we currently have so I expect
      most of the work will be writing those and chasing down bugs.
    * Continue the "cell handling" proposal.


Isabela:
    Last week:
        - working on proposals and reports (never ends)
        - sent Tor Launcher check in email - start to QA the new UI too :)
        - scheduled sync for HS UX tasks (which is schedule for wed)
    This week:
        - send out an earlier reminder to testers for Nov test (left
for last moment the previous months :(
        - review where teams are on roadmap / state of deliverables
        - I have a big deadline on Dec 1st so I might be busy with
that just FIY†

asn
  Last week:
  - Progress with md (#21969) work. #23817 got merged! Might need a bit of a
    fixup still. Progress with #24113 and #23863 but need more work.
  - Reviewed #23100 some more, but got stuck with #23114. Will
continue next week.
  - Wrote two UX improvement proposals for onions:
    https://lists.torproject.org/pipermail/tor-dev/2017-November/012595.html
    https://lists.torproject.org/pipermail/tor-dev/2017-November/012610.html
  - Opened master ticket for onion service DoS improvements: #24298.
    We got many requests to look into this but we didnt really do much,
    let's use that ticket to coordinate.
  - Wrote some things for our censorship circumvention position.
  This week:
  - Hopefully finalize #23817.
  - Attend UX+TB+network team meeting on Wednesday.
  - Take some personal time off.

Mike (may miss part/all of meeting :/)
  Last week:
    -- Git juggled #23100 and #23114 into more reviewable shape.
    - Did some work on #23101
    - Reviewed and discussed #23681
  This week:
    - Maybe get #23101 working?

komlo: (will be on IRC for the meeting ~15 minutes late)
    This week:
        - I want to have a patch for #23881 this week (adding logging
support for rust)

        - One question about this: we use macros on the C side to wrap
log_fn_ (such as log_warn),  I'm not sure if we want to replicate this
in Rust, or if we should instead create helper functions in C.

 - I was thinking about taking a first stap at #23891 (writing a
blogpost about Rust) to get a long headstart and lots of feedback
before the 0.3.3 release


ahf:
    Last week:
        Sponsor 4:
            - Attended the OTF summit together with GeKo, Shari, and the OONI
              people. Some highlights:
              * Met with Hans from the Guardian Project and discussed a lot of
                Android tooling/problems I have had.
              * Met with two activists from Nigeria who gave me some insights in
                mobile usage in their country (focus country for Sponsor 8). We
                exchanged contact information to be able to stay in touch.
              * Conversations with the Briar people about what they could be
                interested in from Tor.
        Sponsor 8:
            - Tried out my `simpleperf` notes in a clean VM to ensure
things worked.
        Misc:
            - Reviewed some small patches in #24167, #24097, and #20963.
    This week:
        Sponsor 8:
            - Talk with Nick about initial set of optimizations for bug #24062.
                - Get `simpleperf` document into the `doc/` tree of tor.git.
                - Begin optimization work.
            - Land my Android logging patch for #24362.
            - Look at Nathan's "split" Orbot work.
               [nickm: please let the rest of us know what you find.]
               [ahf: yes, will update the
https://trac.torproject.org/projects/tor/wiki/org/teams/NetworkTeam/Sponsor8
page for it]
            - Finish automated build's of Orbot-with-tor-master to an fdroid
              repository and find people who are interested in testing it out
              (maybe some people from the network team are up for that?)


Nick:
    - Last week:
        - Worked on separating key/cache directory for limited disk users.
        - Hunted/triaged bugs in 0.3.2 milestone. Fixed some
        - Solicited review on 24107, 24204 APIs (hibernation and
communication for mobile)
        - Started work on making tor restartable in-process (#23847),
which led to #24337 (make all the X_free()s into macros that NULL
their targets).  Is that a good idea?
        - Dove into glibc 2.26 and how to make it play nice with libseccomp2.
        - Backported various fixes.
        - Updated UTF-8 proposal
        - Reviewed David's channels thread.
    - This week:
        - Release 0.3.2.5-alpha (on Tuesday or Wednesday!)
        - Ahf, do you have time to talk about prioritizing the
optimizations to be made for mobile?
          [ahf] Yes! Let's do it early in the week?
        - Taylor, would you like to talk more about bootstrapping some
time? I can make time tonight or tomorrow or maybe even another day.

        [catalyst: yes! tomorrow probably works]

        - Try to finalize fixes for TROVE-2017-{009,010,011}

        [dgoulet] TROVE-2017-12 as well :)

        [nickm] I'm hoping you'll finialize it :) (On the mailing
list) (will look)

        - Try android performance/developer tools recommended by ahf
        - Fix or defer remaining consensus-diff tickets in 0.3.2.x
        - Analyze severity of HS descriptor hashring failures
corresponding to missing ed25519 identities.
        - Continue work on #23847 (restartable in-process tor).
        - Time permitting, start groundwork on making Tor wake up less often.
        - Open review-group-26, if we finish review-group-25 (hint, hint)
        - Time permitting, write a logv variant for komlo to use in
rust for 23881.


catalyst:
    Last week (2017-W46):
        - tried to make more progress on bootstrapping stuff; useful
talk with nickm
        - TB auto-update did some weird things (#23968 is back)
        - useful chat with Sebatian and others on IRC about
expectations about LOG_NOTICE messages and how comprehensible they
should be to relay ops. we should clarify these expectations and
document in easier-to-find places
    This week (2017-W47):
        - talk more with nickm about bootstrapping stuff
        - investigate the cold cache guard state initialization stuff
some more in the client-is-in-the-future scenario (having narrowed
down +3h..+4h as the possibly most problematic case, confirm whether
that holds in practice)
        - review some CoC stuff
        - do we want to move our meeting to 18:00 UTC?

isis:
    last week:
        - fixed some bugs in the moat server and deployed it #22891
        - spent a bunch of time debugging the meek -> apache -> meek
-> bridgedb tunneling, gave up and emailed dcf asking for help
        - started rewriting prop#269
        - reread the XE5 lattice handshake reconciliation paper and
started thinking about what that would look like for another PQ
proposal
        - made my tooling/setup for testing out alpha releases locally better
    this week:
        - finish up proposal writing
        - still haven't done expense reports (oops)


More information about the tor-project mailing list