17:00:25 <nickm> #startmeeting weekly network team meeting
17:00:25 <MeetBot> Meeting started Mon Mar 13 17:00:25 2017 UTC.  The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:25 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:00:27 <nickm> hello everybody!
17:00:32 <ahf> hello!
17:00:44 <haxxpop> hello !
17:00:50 <dgoulet> hi
17:00:59 <nickm> hi ahf, asn, dgoulet, haxxpop, isis, mikeperry, armadev, isabela!
17:01:10 <nickm> let's start with status reports!
17:01:44 <asn> hey hey
17:01:47 <nickm> last week I did some code review and hacking, and I wrote a storage backend for directory stuff
17:01:58 <nickm> now I'm rewriting the directory spool code to not be insane
17:02:12 <nickm> oh. I also worked on consdiff stuff too; fuzzing, testing, etc.
17:02:39 <nickm> once this spool revision is done, I'll work on my code review queue, and work on consdiff integration
17:02:50 <nickm> not clear yet how much of mvdan's remaining integration code I will/won't be able to use
17:03:24 <nickm> I ran into trouble with the storage code: the seccomp2 sandbox requires that every filename be known in advance when the process starts, and that was kind of restrictive.
17:03:33 <nickm> but I think I got a solution that'll work well enough
17:03:52 <nickm> ahf: I hope my storage code and spooling code will turn out to be useful for the new compression stuff
17:04:08 <nickm> Today during discussion I want to talk about:
17:04:13 <nickm> #topic preparations for amsterdam
17:04:16 <ahf> yep
17:04:19 <nickm> #topic how did the triage go
17:04:27 <nickm> End-of-status-report!
17:04:30 <nickm> who's next?
17:04:34 * ahf can go
17:04:38 <nickm> go for it!
17:04:42 <ahf> Hello. Got proposal prop#278 out for review on tor-dev@ and it got some feedback from Nick here on IRC. Cleaned up some I/O code in our utility module that was causing troubles on FreeBSD and OpenBSD (bug #21654). Wrote an implementation of prop#274 in #21641 that is pending review. Started crunching on prop#278 bugs: #21667 and #21663. This week will be doing prop#278 bugs only and I have the bug triage
17:04:48 <ahf> duty. Non-technical stuff: understand the
17:04:50 <ahf> community council setup. That's it :-)
17:06:01 <nickm> ack; sounds nice
17:06:10 <ahf> it was and is nice :-P
17:06:42 <nickm> who's next?
17:06:57 * dgoulet can go
17:07:04 <dgoulet> Last week was *again* mostly #20657 which I've reached the end of circuit implementation. Only descriptor upload logic remains which haxxpop has been helping with.
17:07:06 <dgoulet> Quite happy because this will be ready for AMS work session at least with asn to review the design and choices made in that humongeous branch.
17:07:09 <dgoulet> Good discussion with teor about the IPv4/IPv6 preference for prop224.
17:07:15 <dgoulet> (which ended up with nice code so far ^)
17:07:37 <dgoulet> nickm: and super quick merge_ready this morning :P #21682
17:07:40 <dgoulet> --
17:07:49 <asn> can go next
17:08:13 <asn> i go next:
17:08:16 <asn> Hey guys.
17:08:16 <asn> - During last week I've been very busy with prop224 stuff. I mainly
17:08:16 <asn> worked on the HS ntor branch, and particularly on implementing Python
17:08:16 <asn> integration tests as we have for the normal ntor code. It's pretty much done
17:08:16 <asn> now, I'm just massaging it for the sake of beauty and docs.
17:08:17 <asn> - I also adderssed the #21334 review comments by nickm, which got me knees-deep
17:08:17 <asn> into #21693. #21693 is a pretty messy but important ticket and I hope to have a
17:08:19 <asn> draft ready this week.
17:08:19 <asn> - I also worked on #21415 aka The Guard Bug, which I debugged and hopefully fixed
17:08:21 <asn> for good.
17:08:21 <asn> - I also did an initial review of #21039. Finally, I did some spec updates for
17:08:23 <asn> prop224 and reviewed a bit of haxxpop's code. Pretty busy week but I had
17:08:23 <asn> nothing else to do while recovering from tonsilitis so...
17:08:26 <asn> EOF
17:08:40 <nickm> dgoulet: thanks for the review!
17:09:04 <nickm> asn: can I merge the HSDesc branch soon, or should I wait for the base64 fix?
17:09:16 <asn> good question.
17:09:22 <asn> i think the hsdesc branch is ready to be merged
17:09:32 <nickm> personally I'd rather take the branch, and then merge any additional fixes to master
17:09:32 <asn> i'm not sure when i will have a base64 fix that is _ready_ to be merged
17:09:35 <nickm> ok
17:09:48 <asn> and when I have one, it will likely based on the hsdesc branch
17:09:52 <asn> so i think merging it is a good idea
17:09:56 <nickm> will do!
17:09:59 <asn> ack thx
17:10:26 <nickm> dgoulet, asn: Also, would one of you be interested in setting up fuzzers for this format? :)
17:10:30 <haxxpop> As dgoulet and asn said, I did the responsible HSDirs stuff but there is some more work to do :)
17:10:50 <dgoulet> nickm: teor opened a ticket for that and yes we need to do that, maybe a little workshop in AMS? :)
17:10:52 <asn> nickm: yep we should probably do that!
17:11:02 <nickm> keen
17:11:36 <nickm> anyone else?  isis mikeperry armadev  isabela ?
17:11:40 <isabela> yo
17:11:44 <nickm> hi isabela !
17:12:12 <isabela> i am still living the reports/proposals life :) but this week i will try to help yall organize for amsterdam
17:12:17 <isabela> done o/
17:12:29 <mikeperry> I am here, but again I am working on Firefox code review. I also have a talk at the IETF right after the Tor meeting, so I need to work on that before I go, too (which means this week)
17:13:10 <asn> isabela: just to make sure, we are good with sponsorr task 3.1, right?
17:13:16 <isabela> yes
17:13:20 <asn> isabela: the moment i learn that task 3.3 is done, i will send a similar email.
17:13:29 <isabela> sounds perfect
17:13:33 <asn> most excellent
17:14:33 <nickm> ok. any more updates, or time for discussion?
17:16:17 <nickm> time for discussion!
17:16:59 <nickm> so, last week we did some initial triage on 031.  I kicked out every ticket T such that the priority of T was low, and T had no owner, and T had no sponsor, and T's points were >=.5
17:17:08 <nickm> That leaves 120 tickets or so
17:17:36 <nickm> I didn't remove anything in needs_review, needs_revision, or needs_information.  Maybe we should have a look at stalled needs_revision tickets and see if we want to pick any up, or postpone them
17:17:40 <nickm> any other triage thoughts?
17:18:49 * isabela looks at it one sec
17:18:51 <nickm> Also, we have a problem with "Tor: Unspecified": It's huuuuuge
17:19:03 <dgoulet> I see a lot of tickets flagged for proposal that I assume we want in 031
17:19:12 <dgoulet> so filtering them out could be a start
17:19:21 <dgoulet> a start for triage
17:21:08 <isabela> dgoulet: which proposal number you are looking at?
17:21:17 <dgoulet> 87 tickets if I filter out prop224, prop140 and prop278
17:22:17 <isabela> nickm: i would say tickets that tickets that are not need review, need revision, mergy ready and has no sponsor or owner should be considered to be moved out
17:22:18 <nickm> so you mean, fix no bugs unless they're proposal-related? It's a start... but there are more things we want to do as well, yeah?
17:22:39 <isabela> i think i see 21 tickets like that
17:22:40 <nickm> isabela: I did that except for stuff that had "very high" priority or very low points
17:22:45 <nickm> isabela: (I think)
17:22:54 <dgoulet> no! not move those 87 tickets out, I meant start with triage on those 87 tickets since the others we'll keep them
17:24:03 <armadev> hello i'm here. just realized i dropped a phone call with a funder earlier today, so scrambling on that.
17:24:23 <asn> dgoulet: makes sense
17:24:24 <asn> i
17:24:39 <asn> dgoulet: got a query for those 87 tix?
17:24:39 <isabela> nickm: gotcha - i am seeing some low priority still. like #19669
17:24:52 <dgoulet> https://trac.torproject.org/projects/tor/query?status=accepted&status=assigned&status=merge_ready&status=needs_information&status=needs_review&status=needs_revision&status=new&status=reopened&keywords=!~prop224&summary=!~prop140&summary=!~Prop278&summary=!~prop278&milestone=^Tor%3A+0.3.1.x-final&group=status&max=150&col=id&col=summary&col=milestone&col=status&col=owner&col=type&col=priority&col=component&
17:24:52 <dgoulet> desc=1&order=id
17:24:52 <asn> i feel tempted to handle some now, but a spreadshheet approach is probably wiser?
17:24:54 <isabela> nickm: by low priority i mean - someone marked it like that
17:24:55 <dgoulet> that is the big one I have
17:25:06 <dgoulet> prop140 and prop278 don,t have keyword so you have to filter out by Summary :(
17:25:13 <nickm> isabela: that one is in needs_information
17:25:15 <dgoulet> asn: yeah spreadsheet++
17:25:29 <isabela> nickm: i was counting needs_info too
17:25:33 <nickm> ok. anybody want to set up the spreadsheet ?
17:25:50 <isabela> sure
17:26:21 <isabela> is it ok if gets done wed tho? i have other work i have to prioritize today and tomorrow
17:26:38 <isabela> or do you mean just a dump?
17:26:42 <isabela> not organize things on it
17:26:50 <ahf> i'll add prop278 keywords to tickets missing it :/
17:26:58 <dgoulet> ahf: neat, you can do a batch modify
17:27:03 <nickm> just a dump would be fine IMO
17:27:28 <isabela> nickm: cool i will do it now
17:28:01 <dgoulet> asn: we kind of have to decide if those are realistic for 031: #20699, #20594, #18098
17:28:10 <dgoulet> asn: keyword "prop224-extra"
17:28:27 <asn> im pretty sure it's not realistic
17:28:32 <dgoulet> control port is ... huge
17:28:35 <asn> i wouldn't feel bad about removing those right now
17:28:40 <dgoulet> ok moving them to 032
17:28:43 <asn> and _if_ we ever have extra time we can readd them
17:29:00 <asn> ack
17:29:09 <dgoulet> blam
17:29:14 <dgoulet> 3 out of 031
17:29:18 <asn> kablam
17:30:34 <asn> so should we reserve a slot for real life triaging in amsterdam as well?
17:30:41 <dgoulet> ahf: for the record I like having "prop278:" in the title :) but keyword is very useful for searches
17:30:46 <asn> or should we just figure out the right time when we get there? :/
17:31:11 <ahf> dgoulet: thanks for the suggestion with batch modify - that was useful. done now.
17:32:04 <isabela> nickm: https://docs.google.com/spreadsheets/d/1UTaLO2na4-lcxbPSeYtugkd_qqaUCO82cnSFkYt8d6w/edit?usp=sharing
17:32:48 <nickm> thanks!
17:32:54 <nickm> any more on triage, or ready for next topic?
17:33:27 <asn> i guess... what do we do wit hthe spreadsheed?
17:33:36 <asn> should we all check it out by some day?
17:33:55 <nickm> isabela:  ^ ?
17:34:22 <asn> like should i take a look tomorow or the day after, and color background red the tickets i consider movable to 032?
17:34:38 <dgoulet> (ok all prop140 ticket have a prop140 keyword now)
17:35:15 <isabela> what i did before is that i would organize other sheets for instance by person or sponsor and everything else we are absolute sure we want to do it for 031 (is a must do)
17:35:18 <asn> or should we prepare the spreadsheet so that it's a guide for amsterdam?
17:35:33 <isabela> do a quick capacity calculation / reality check
17:35:58 <nickm> like, make sure points are right -looking, see how many we have, decide what to do?
17:35:59 <isabela> and list what is out of that cut for triage - cuz till now we tried to remove as much as we could
17:36:25 <dgoulet> (btw this point thing is weird because the maximum is "6" but for service implementation have been spending weeks on it...)
17:36:43 <asn> :)
17:36:45 <dgoulet> so at best I can know how much bugs I can take that aren,t prop224
17:36:58 <asn> dgoulet: i guess we should have splitted it into more tickets? ...
17:37:00 <isabela> dgoulet: when that happens it means the ticket should have been broken down into multiple tickets
17:37:10 <dgoulet> well that doesn,t work super well with developmetn :(
17:37:17 <asn> yep true
17:37:24 <dgoulet> multiple branches and that feature would have been an horror story
17:37:50 <asn> maybe not one branch per ticket is required. but i could have made a branch for ntor, and another one for time period, etc.
17:37:51 <dgoulet> so we broke it down in the commit and asn did for instance many bits of it which I put in the main branch of that ticket
17:38:07 <dgoulet> asn: I guess but then why the ticket? keeping time ?
17:38:22 <asn> dgoulet: i guess so......
17:38:33 <isabela> this is for avoiding over committment
17:39:02 <isabela> we have to be able to somehow say how much work one can do over a period of time
17:39:04 <asn> dgoulet: also it would have been pretty genius of us to predict all the various bits of service-side beforehand...
17:39:15 <isabela> and be realistic as possible on it
17:39:17 <asn> dgoulet: and by genius i mean hard. or lots of time.
17:39:23 <dgoulet> yeah
17:39:54 <dgoulet> isabela: now we do have a much better picture of what that whole things is so maybe I can open the tickets but they will almost all be with "6" points
17:40:02 <dgoulet> and I'll end up having 50+ points alone
17:40:13 <dgoulet> what do you think?
17:40:55 <dgoulet> (and asn 50+ points also :P)
17:41:11 <isabela> i would suggest to think more like how many points i do per week
17:42:05 <isabela> than you can say ok i have 4 tickets of 6points each does that fit in a month of work? can I take that? or should i take 2 now and the other 2 next month
17:44:16 <dgoulet> hrmm ok
17:44:28 <asn> oh man that's so hard... i just take everything there is and charge at it.
17:44:33 <asn> and what remains spillsover to the next month.
17:45:05 <dgoulet> yeah
17:45:20 <dgoulet> that development is kind of liner of "We have to make it happen and we adapt as we go"
17:46:05 <asn> like in january i could never predict that dgoulet would need me to make a time period branch (6 points) in february
17:46:20 <asn> (just example)
17:46:21 <isabela> right now how you are determing that you can do all the tickets you are taking for 031?
17:46:48 <dgoulet> isabela: no way to tell, we just charge for prop224 to be in 031
17:46:53 <dgoulet> and we think we can pull it off :)
17:46:56 <isabela> do you nkow if they will take you 3 months of work, and are you counting the fact that you will split your time with other stuff like bug triage and code review
17:47:06 <isabela> dgoulet: hehe
17:47:08 <dgoulet> (well actually for the sake of our sponsor happyness, we need to)
17:47:38 <isabela> so yeah, that is the goal of points to avoid being in a situation where you might be over committing
17:48:00 <isabela> dgoulet: i understand that, and sometimes you cant get away from it
17:48:08 <dgoulet> yeah seems like an adaptative system basically
17:48:24 <dgoulet> we update as we go our prediction
17:48:45 <isabela> but for a sane happy life :) we should try to have a sense of what is a reasonable amount of work one can do and plan our work with that capacity in mind and plan deliverables for sponsors within that
17:49:31 <dgoulet> isabela: yes makes sense. Ok so now that we have *not* done that for #20657, should I do it? :)
17:49:40 <isabela> lol
17:49:46 <dgoulet> (break it down in multiple tickest with points?)
17:50:10 <isabela> giving what you said above - is a do or die situation :P
17:50:16 <dgoulet> kind of ahaha
17:50:20 <isabela> i would suggest that you and asn does not take anything else
17:50:22 <dgoulet> and many promesses were made for 031 :D
17:50:54 <isabela> so more important of adding points in this situation in particular is to make sure you and asn dont take anything else
17:50:57 <isabela> so you can meet it
17:51:04 <nickm> should we do a little on our "getting ready for AMS" topic before we run out of meeting time?
17:51:10 <isabela> because i bet it will be already over committing
17:51:20 <dgoulet> nickm: woa meeting.. damn sorry
17:51:21 <isabela> nickm: sorry
17:51:23 <isabela> :(
17:51:28 <nickm> my current prep plan is "avoid working Friday through Monday" :)
17:51:28 <dgoulet> isabela: ok makes sense
17:51:37 <dgoulet> wise
17:51:37 <nickm> but please make sure I know about anything I should do to get ready?
17:51:47 <isabela> :) ok
17:53:21 <dgoulet> ok quick note on AMS
17:53:27 <asn> prop224 wont see me cry ; we gonna make it ; its do or die
17:54:09 <dgoulet> me and asn have scheduled a prop224 work session which is on the 23rd at 11am, not mandatory ofc but the point is to discuss design and code structure of what we have now in development
17:54:15 <dgoulet> and plan the rest basically for 031
17:54:37 <dgoulet> anyone welcome to attend but won't be a "prop224 informative session" :), very narrow objective
17:54:56 <dgoulet> "Room 2" wherever that is
17:54:57 <nickm> is there a master calendar for all this somwehere?
17:55:02 <dgoulet> https://storm.torproject.org/shared/_xuQsyM7Lkssjv9gKLw2rePMn9JYhtUBx1nr5WgwiU7
17:55:03 <isabela> yes
17:55:05 <nickm> on the wiki, I'd bet?
17:55:07 <nickm> cool
17:55:18 <dgoulet> which also have our slot for our net team meeting at 16h
17:55:26 <dgoulet> and then at 14h30! :)
17:55:42 <dgoulet> UX/App/Net team join effort for some discussion about prop224 integration
17:55:45 <dgoulet> so many meetings
17:55:55 <isabela> hehe
17:56:07 <dgoulet> (and it could also be about other things than HS also if needed/time)
17:56:34 <dgoulet> oh and my name being there doesn,t mean I moderate lol
17:56:48 <asn> suggestion: we all look at the triaging spreadsheet this week and have it ready for ams. and then in ams, we arrange a time to go over spreadsheet and do the final triaging?
17:56:55 <nickm> also I tried to get jon to organize a net-team dinner.  please get back to him if you haven't yet!
17:57:14 <dgoulet> nickm: Shari did sent a confirmation about that I believe
17:57:22 <ahf> is there a spreadsheet like that for each day at the event somewhere? don't see it on the wiki page on trac
17:57:34 <asn> have others received email about dinner? i haven't but maybe stuck in my spam filters. riseup is doing this all the time lately.
17:57:44 <isabela> ahf: not that i know
17:57:48 <ahf> ok
17:57:55 <dgoulet> asn: seems shari CCed me and vegas-team@ lol
17:57:58 <ahf> asn: i got a confirmation from jon that he received my email - that is all
17:58:02 <isabela> dgoulet: hehe
17:58:10 <isabela> dgoulet: fwd to network team list?
17:58:17 <dgoulet> yeah fwd, nothing sekrit in there
17:58:23 <asn> kk
17:59:02 <nickm> okay. isabela: should I wait on more info from you about stuff to do before/at the meeting?
17:59:11 <nickm> even a simple outline of "make sure this gets done" would be super helpful
17:59:13 <dgoulet> tl;dr;, 18h00 on the 23rd is our dinner reservation :)
17:59:19 <nickm> cool
17:59:21 <isabela> nickm: ok o/
17:59:23 <isabela> will do
17:59:42 <asn> dgoulet: excellent'
17:59:52 <ahf> dgoulet: oh, this is the meet-and-greet dinner?
17:59:59 <dgoulet> ahf: no it's network team dinner :)
18:00:05 <ahf> then it's the same day?
18:00:10 <dgoulet> oh is it lol ?
18:00:11 <ahf> If you plan on attending the "meet and greet" dinner on Thursday, March 23rd, please add yourself to this table:
18:00:12 <isabela> !
18:00:14 <isabela> ai ai ai
18:00:20 <asn> hehe
18:00:24 <dgoulet> ok there is some confusion here... ahaha
18:00:27 <isabela> hehehehehe
18:00:42 * dgoulet emailing jon
18:00:44 * ahf really don't want to miss either of those two dinners :-S
18:00:52 <isabela> dgoulet: tx!
18:00:54 <dgoulet> for sure they can't be at the same time ;)
18:00:59 <isabela> hehehe
18:01:07 <asn> they are isolating us!!!
18:01:09 <asn> j/k
18:01:20 * isabela puts on foil hat
18:01:28 <ahf> we just get a table for ourself in an inflatable plastic bubble
18:01:53 <dgoulet> children's table
18:02:09 <nickm> bad kids' table ;)
18:02:11 <isabela> hahaha
18:02:16 <nickm> ok, any more for today?
18:02:21 <ahf> lol dgoulet
18:02:31 <nickm> else I will #endmeeting and we can all get back to haxxing.
18:02:35 * dgoulet is good
18:02:41 <isabela> o/
18:02:47 <isabela> good too
18:02:56 <asn> good here :) let it go!
18:02:58 <nickm> oh!  Also everybody plrease remember to make everybody, especially new people and shy people, welcome next week.
18:03:01 <nickm> #endmeeting