[tor-project] Network team meeting notes, 20 Feb 2018
Nick Mathewson
nickm at torproject.org
Wed Feb 21 00:31:42 UTC 2018
Hi, all!
You can find the proceedings of our latest meeting at
http://meetbot.debian.net/tor-meeting/2018/tor-meeting.2018-02-20-17.59.html
Our notes are below:
==================
Network team meeting pad, 20 Feb 2018
Welcome to our meeting! Tuesday (normally Mondays) at 1800 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 previous weeks:
* https://lists.torproject.org/pipermail/tor-project/2018-January/001644.html
* https://lists.torproject.org/pipermail/tor-project/2018-January/001647.html
* https://lists.torproject.org/pipermail/tor-project/2018-February/001654.html
* https://lists.torproject.org/pipermail/tor-project/2018-February/001664.html
Old Announcements:
* 0.3.3.x is in feature-freeze. No new features (except for
#24902, which has permission.).
* 0.3.3.2-alpha came out Saturday 10 Feb.
* Please, work on bugfixes! It would be great to have this release
release on time. (Planned date is 15 April)
* There are tons of tiny 0.3.3.x tickets. Maybe if you can do one
of them in <X minutes for some small X, you should just do it?
* We moved the hackfest 2 days later (Wednesday 30th - Sunday 3rd)
to avoid clashing with Memorial Day
dgoulet, teor, and isabela are organising it off-list.
See the network-team list for updates.
* No GSoC for Tor :'( [this time]
Announcements:
* Jon is booking accommodation for the May hackfest.
* The 0.3.4.x merge window opened on Feb 15.
* Review-group-32 is open! There are 2 tickets assigned to teor
and mikeperry to review.
Maybe we can move on to review-group-33?
https://trac.torproject.org/projects/tor/query?status=!closed&keywords=~review-group-32&group=status
* Nick just made a review-group-33. Please take tickets to review.
* Please fill in the code subsystems spreadsheet at
https://docs.google.com/spreadsheets/d/1Ufrun1khEo5Cwd6OwngERn829wU3W3eskdrriaYfUBQ/edit#gid=1974867547
If you review tickets, feel free to add a column with your name.
* Team rotation roles was empty last week and this week, ahf have
taken this week, but its empty for next week:
https://trac.torproject.org/projects/tor/wiki/org/teams/NetworkTeam/TeamRotations
* isis cannot attend Rome :(
* prop#245 + prop#282 meeting today (tuesday) at 23:00 UTC
Discussion:
- Topics
- Can we remove the *.nmake files in tor.git? I have a feeling
that they are unmaintained/unused?
- Is all the dates final for the May hackfest? (would like to
order tickets early from Europe while they are cheap).
- Isabela's heat-mapping exercise with areas of tor work: what are
our priorities? https://pad.riseup.net/p/38VbTZUIzG4X
======
teor:
* Last week:
- Implemented and tested more Experimental PrivCount circuit
and onion service stats
(1.5 features left, and then it's feature-complete. I hope!)
- More May hackfest planning
- More Proposal meetings
- Detailed design and numerical analysis for uniform random in
Rust, but still no code
- Reviewed a few tickets, merged a chutney network for bwscanner
* This week:
- Final Experimental PrivCount features (no, really, this week)
- Eventually make a trac user page that lists the tickets I'm
working on in priority order.
Add something similar to the network team queries page.
- I should get time to review or revise more tickets later this week
catalyst:
last week (2018-W07):
- reviewed #25185 (Rust cstr stuff)
- learned more Rust stuff, including some doc-test things
- patch for #21074 check-changes
- travel logistics for Rome meeting and Seattle hackfest
- CoC and SoV feedback
- helped pastly with IRC moderation stuff
this week (2018-W08):
- work on 0.3.3 bugs
- review more tickets
- help out with floating point doom
- update roadmappy stuff
- more travel logistics
- monitor CoC and SoV feedback
dgoulet:
* Last week:
- We got #24902 backported \o/
- Discussion with armadev on #25226 for which it could be that we want a
hard limit of cell queue size per circuit.
- Did a branch for #17945 as a possible solution. Need feedback.
- Opened and worked on #25268 to remove the cmux round-robin algorithm.
- I've also started working on a branch to improve our fast path (per cell)
within the cmux subsystem. It touches also #25152. Not ready for merge or
anything, it was more experimental to see how that would play out.
- Worked with asn on the subsystem maintainer sheet. *PLEASE FILL IT UP*
- A little bit of May/June Hackfest logistics with Jon.
- Was quite a week for bad relays so spent some time there.
- Helping atagar with some HSv3 questions and examples for Stem integration
- Sent to tor-dev@ my 034 enhancement plans:
https://lists.torproject.org/pipermail/tor-dev/2018-February/012949.html
- Some ticket triage including 033-must tagging. See timeline:
https://trac.torproject.org/projects/tor/timeline?from=Feb+20%2C+2018&daysback=8&authors=dgoulet&tag_query=&milestone=on&ticket=on&ticket_details=on&changeset=on&wiki=on&tags=on&sfp_email=&sfph_mail=&update=Update
* This week:
- Hope to work on 034 enhancement so we can get them early in the merge
window. Especially the scheduler stuff.
- 033 tickets are still a reality.
- Get #24767 in a merge_ready state (asn did a first review)
Nick:
* Last week:
- Too many meetings ;)
- Worked on TROVE fixes, rust protover fixes, etc. (See
network-team-security, #25250, #25251)
- Opened 0.3.4, merged much of backlog.
- Administrative stuff for roadmapping/planning
- Conferred with isabela about roadmapping etc.
* This week:
- Pre-announce next week's security releases
- Prep for next week's security releases
- Review and merge more.
- Finish 0.3.3 fixup triage [033-must, continued]
- Write up an implementation plan for improved hibernation
backend, based on
https://gist.github.com/ahf/d22a4a98abe4651d07181a993cebdd21
[continued]
- Get an official repo for tor-rust-dependencies
- Get an official repo for tracking "latest stable", "latest alpha", etc
asn:
Last week:
- Helped with roadmapping. Revived subsystem maintainer spreadsheet.
Please fill it out if you haven't:
https://docs.google.com/spreadsheets/d/1Ufrun1khEo5Cwd6OwngERn829wU3W3eskdrriaYfUBQ/edit#gid=1974867547
- Triaged tor-hs and tor-guard tickets of 033.
- Arranging Tor seminar at ENS Paris at the end of February with the
help of maker.
- Reviewed #24767.
- Reviewed #24343.
- Reviewed #21425.
This week:
- More roadmapping fun.
- More reviews/bugfixes.
- Think about HS statistics that we could gather (reply to teor's email).
Mike:
Last week:
- Reorganizing and refactoring vanguards repo
- Reviewed #22926
- Reviewed guardraction code - are we sure we want to rip it out
rather than just disable torrc?
[asn: i think we want to rip it off to simplify the codebase.
otherwise we could keep it lingering on until and if we have firepower
to fix it...] [mike: do we believe that a new from-scratch
implementation will be simpler?] [asn: no. i think we just realize we
wont ever have enough time to fix it and maintain it] [mike: hrmm..
that seems defeatist and potentially wasteful :/] [asn: might be. it's
just that the code has been sitting around for many releases and it's
been broken and no one wants to fix it or maintain it. also dirauth
ops are complaining about the python script and we should tell them to
kill it] [mike: it seems like disabling and deprecating the python
script can be separate from removing all of the functionality from
tor-core... is this the type of thing that will always be easier to do
in an external program, because of statekeeping?] [asn: yes it's
because of statekeeping that we do it as external. and i agree that
perhap disabling the python script is separate from removing code from
tor. as an argument for keeping it, the guardfraction tor code is
fairly isolated and abstracted from the rest of the code]
ahf:
Last week:
Sponsor 8:
- Walked over S8 roadmap tasks.
- Baselined some memory usage experiments with Shadow+dtrace that
was done with tor+simpleperf before.
- Made a Tor/Orbot APK where Tor stores its log on the device and
the device periodically submit some grep'ed lines to a web
service.
Misc:
- Reviewed: #25095
- Went over the different documents discussed at last
meeting + the subsystem survey.
- Discussed some PT enhancements after the PT meeting in #tor-dev.
- Played around with Windows API's and cross compilation
of Tor for Windows.
This week:
Sponsor 8:
- Baseline event loop statistics and all the client tests in Shadow
Misc:
- Look into #24368 (zstd tuning).
- Help out with some bugfixing.
- Next weekend I hope to get a prototype/demo running with a PT
subprocess that can do logging via tor.
isis:
last week:
- finished moat work #24432
- dumped all the unallocated bridges into the new moat
distributor to kickstart it #25246
- reviewed #25150 and #24658 and #25071
- tiny fix for duplicate header inclusion #25271
- began making rust wrappers for our C digest code #24659… can
i include a new dependency on the digest crate?
- researched the safety of zero-sized types accross FFI boundaries
this week:
- documenting moat reflector setup somehow as much as possible?
- continuing large create cells #24986 impl
- continuing rust digest wrappers #24659
- continue scheduling proposal meetups and have meetings
- meeting this week with teor about prop#245 and prop#282 (and
whomever else would like to join) (at 23:00 UTC today)
isabela:
last week:
- started the roadmapping process - today will talk about
voting on priorities
- met nick o/
- met roger o/
this week:
- continuation of roadmapping process - organize first draft
of tasks for 6 months roadmap
- send out feb network testing emails
More information about the tor-project
mailing list