[tor-project] Network team meeting notes from 27 Aug

Nick Mathewson nickm at torproject.org
Tue Aug 28 23:34:49 UTC 2018


You can find the logs at

Below are our notes from the meeting.

= Network team meeting pad! =

Welcome to our meeting! Mondays 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.)

After each week's meetings, the contents of this pad will be sent to
tor-project @ lists.torproject.org.  After that is done, the pad can
be used for the next week.

== Previous notes ==
11 June: https://lists.torproject.org/pipermail/tor-project/2018-June/001828.html
18 June: https://lists.torproject.org/pipermail/tor-project/2018-June/001835.html
25 June: https://lists.torproject.org/pipermail/tor-project/2018-July/001863.html
2 July: https://lists.torproject.org/pipermail/tor-project/2018-July/001866.html
9 July: https://lists.torproject.org/pipermail/tor-project/2018-July/001884.html
16 July: https://lists.torproject.org/pipermail/tor-project/2018-July/001888.html
23 July: https://lists.torproject.org/pipermail/tor-project/2018-July/001926.html
30 July: https://lists.torproject.org/pipermail/tor-project/2018-July/001928.html
6 Aug: https://lists.torproject.org/pipermail/tor-project/2018-August/001936.html
13 Aug: https://lists.torproject.org/pipermail/tor-project/2018-August/001947.html
20 Aug: https://lists.torproject.org/pipermail/tor-project/2018-August/001954.html

== Stuff to do every week =

* Let's check and update the roadmap.  What's done, and what's coming up?

url to roadmap:

* Check reviewer assignments at

* Also, let's check for things we need update on our spreadsheet!  Are
there important documents we should link to?  Things we should

* Check rotations at

* Community guides, it's time to hand off to the next guide!

* Let's look at proposed tickets!  [but see discussion]

== Reminders ==

* Remember to "/me status: foo" at least once daily.
* Remember that our current code reviews should be done by end-of-week.
* Make sure you are in touch with everybody with whom you are doing
work for the next releases

---- 27 August 2018

== Announcements ==

Sponsor8 ends on Dec 31, and we have some deliverables that needs to
make 035.  The 035 merge window closes on Sept 15.  With that in mind,
we need to prioritize these deliverables, and ask some folks to step
out from rotations and most code reviews for the next month or so, so
we can get it all done.

Code reviews will also prioritize sponsor8 work, so, some reviews
might be delayed and not added to the weekly review spreadsheet.

For the rest of us, we don't intend that reviews and rotations should
take up your whole life.  If you're spending more than 20% or so of
your week on reviews and rotations, then we might need to recalibrate.
Talk to Nick and Isa if that sounds like the situation you're in.

We will resume normal team operations once we get the sponsor8
deliverables done :) thank you for the understanding.

Sbws can match torflow!
Here are our current sbws design questions:

We fixed some failures in chutney, tor 0.3.4 and tor master.
Please pull the latest chutney, and base new branches on the latest tor.

== Discussion ==

Who do we want at the teams day at the Mexico City meeting?
(teor suggests: staff, and people who are making significant
contributions to core tor, sbws, etc.)
Have we missed anyone? Is it too late to ask?

- Alex from the Rust team is planning to join, but noting him here to
make sure we send an official invite

Accepting PRs not on Github: [komlo]

- It is hard to review PRs that don't have Travis output- how do we
feel about requiring PRs on github for external contributors so that
we can see the status of Travis builds? Is there a non-github
alternative where we can still see Travis output? Is this something
that should be in our Coding Standards? [catalyst: i'm inclined to
agree. PRs on our GitHub repo also run Appveyor and Coveralls CI,
which not everyone has enabled on their forks. in practice, when i
review a non-GitHub patch or contribution, i usually push a version to
my fork and make a PR so that CI runs.][Asking for PRs and/or creating
PRs as needed sounds like it can be a community hero thing. -nm][+1

== Updates ==

    Last week:
        - Got NSS working.  It now passes all its tests, and a chutney
network bootstraps, and a client connects to the network. Going to add
a bunch more tests, and hunt down a couple of memory leaks, but I
think we've got a good start here. Timeline for review? [catalyst: i
can hand off to ahf if he's willing]
        - Helped diagnose timing issues in voting w mixed chutney
networks (27146)
        - Looked at memory profiles, and opened subtickets &
subsubtickets for #26630.  (We're wasting a lot of memory in our
directory code.)
        - Helped diagnose bridge bootstrapping issues w chutney on 0.3.4 (27080)
        - Reviewed and merged a bunch of stuff
        - Released New release date for How about
September 10? [works for me -teor]
        - Bug triage. Cleared out tickets in "- Select a component"
and tickets without a milestone; worked on incoming tickets.
        - Fixed a crash caused by new openssl not liking some of the
tricks in test_tortls.c (#27226)
    This week:
        - Actually write the sidechannels whitepaper that I said I'd
finish last week :/
        - NSS: add a bunch more tests, and hunt down a couple of memory leaks
        - NSS, Privcount, etc: revisions based on reviews, if any
        - Work on one or more memory reduction tickets under #26630 and #27243.
        - Taking off Thursday.
        - More review and merging

teor (offline):

   last week:
       - sbws can match torflow's results (next step: design feedback)
         Thanks to juga and pastly for all their hard work!
       - Fixed broken bridge tests in chutney (#27080)
       - We changed consensus timings between 0.3.3 and 0.3.4 (#27146)
       - Helped with Rust linking (#27272, #27273, #27274)
       - Other fast fixes, reviews, and ticket triage
       - Helped purge old trac admin accounts

   this week:
       - Review the fixes to the vanguards patch (#25573)
       - Get chuntney working in mixed 0.3.3/0.3.4 networks (#27146)
       - maybe we should run chutney in travis CI? (#20647)
       - our macOS time API seems to return weird results on i386 (#26987)
       - Create an updated list of fallback directory mirrors (#24786)
       - PrivCount in Tor: blinding and encryption (#22898: reviews,
fixes, design meeting)
       - bwauth work (#27135)

  Last week:
    - Worked on #27246 for Sponsor 8.
    - Worked on #27215 which makes the HS default version to 3 in 0.3.5.
    - Memory profiling of Tor client. From the results, nickm opened #27243 and
      the childs for improvements. For the curious:
    - Reviewed HS client authorization branch (#20700). Almost done after the
      on-disk design now in the spec. Hoping to have a potential merge
      candidate this week.
    - Reviewed a series of tickets (see Timeline).
  This week:
    - Keeping myself available for #20700 next revision (client HS auth).
    - Review anything related to S8 tor client memory optimization.
    - Possibly also help asn on the HS open bugs if time permit.
    - Work on prop289 if time allows it.

 Last week:
 - Wrote patch for #26980. This was a reliability issue for HSv3 because it was
   causing descriptor upload failures. Patch has been merged, and my HSv3 has
   been functioning well ever since.
 - Review #23576, #13843, #24880, #22958.
 - Was away from Monday and up to Wednesday.
 This week:
 - Start working on "onion services open issues" blog post with Mike. My plan
   is to start by writing an early draft, and then pass it to Mike for
   feedback. Rinse and repeat.
 - Get client auth (closer to) merged.
 - Try to get some time to seriously look into #25882. This is the main
   reliability issue we have in HSes (v2/v3) right now, and it's quite
   fundamental because it's caused by simple circuit timeouts. Diagnosing and
   fixing this issue is probably some work since it requires a stressed out HS.


Last week:

- Returned to Tor after some vacation.

This week:

- Sponsor 8:

- Get back on track with #25502 (PT error reporting)

- Figure out where we are with memory optimization.

- Look into Nick's NSS work.

- Misc:

- Go over review backlog.

komlo (offline):
    - Last week:
        - Rough draft of a peer reivew checklist for research on Tor,
will finish this in a month after I can meet with Ian to discuss. Once
we have this I'll pair with Mike to finish this up and publish
        - RustConf catch up (sent an email to get Alex up to speed so
he can help with diagnosing build issues)
    - This week:
        - Ad-hoc Rust ticket review, there are a lot of small tickets
that need review

haxxpop (offline):
    - Last week:
        - Discussed the client auth code with dgoulet
    - This week:
        - Revise the client auth pr according to the comments

    - last week (2018-W34):
        - ran into #27226 on macos. reviewed Nick's fix
        - made some progress on #22266 (bootstrap progress reporting)
    - this week (2018-W35):
        - community guide rotation [to nickm]
        - more on #25502 and #22266 (bootstrap progress reporting)
        - more home maintenance backlog

  Last week:
      - Updated #25573 (half-closed connection id tracking) based on review
      - Kept digging into #27066. Have a potential stopgap fix.
      - Rebased WTF-PAD branch onto master
  This week:
      - Look over and reply to sbw mail
      - Continue work on WTF-PAD as per

More information about the tor-project mailing list