14:28:42 <karsten> #startmeeting metrics team meeting
14:28:42 <MeetBot> Meeting started Mon Sep 10 14:28:42 2018 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:28:42 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:28:44 <karsten> hi!
14:28:47 <irl> hi!
14:28:49 <karsten> finally, we have a meeting.
14:28:52 <irl> yes
14:29:20 <karsten> https://storm.torproject.org/shared/5h1Goax5eNusxjXJ_Ty5Wl7hFR1uqCReUiN8xdlBG8T <- agenda pad
14:29:37 <karsten> and we have a few things on the agenda.
14:29:56 <karsten> ready to start, or do you want to add more?
14:30:08 <irl> i don't have anything else to add
14:30:11 <karsten> ok.
14:30:15 <karsten> * Travis CI for Metrics codebases (irl)
14:30:40 <irl> this was something i was hacking on before i left for the last round of travel
14:30:49 <karsten> did you know we have this? https://jenkins.torproject.org/job/metrics-lib-master/
14:31:05 <irl> i did not
14:31:23 <irl> the main thing i wanted to add was automatic static code analysis and coverage reporting
14:31:25 <karsten> should we try to get more code bases into jenkins?
14:31:28 <irl> does jenkins do this?
14:31:33 <karsten> hmmm.
14:32:03 <karsten> I don't really know what jenkins can do.
14:32:24 <karsten> where would the travis stuff run?
14:32:31 <irl> i didn't actually get either of those things working though, only running the test suite so far
14:32:43 <irl> travis runs on the free travis ci infrastructure that is integrated with github
14:33:28 <karsten> ok. so, I can see the value in doing this.
14:33:42 <karsten> should we involve hiro in this?
14:33:52 <irl> perhaps
14:33:58 <karsten> or is that not necessary, because things run on github machines anyway?
14:34:28 <irl> we would need to have things configured on the github side, which i think only github people can do
14:34:31 <irl> and hiro is one of those people
14:34:39 <karsten> ok.
14:34:42 <irl> maybe i am one of those people now that i've been doing git things
14:34:48 <irl> it's not entirely clear to me
14:35:12 <irl> i will follow up with hiro and see
14:35:18 <karsten> ok.
14:35:29 <karsten> so, we haven't paid as much attention to jenkins in the past.
14:35:38 <karsten> like, we didn't notice it was broken for months.
14:35:43 <irl> oops
14:35:46 <karsten> but I wouldn't say that is jenkins' fault.
14:35:59 <karsten> of course, having some sort of notification (email) would be good.
14:36:09 <irl> it may be possible to get jenkins to build on user branches, which would be really useful
14:36:16 <irl> once master is broken it's probably too late
14:36:24 <karsten> (the issue was that we didn't update libs. no bug in the code.)
14:36:33 <irl> right
14:36:49 <karsten> makes sense, re user branches.
14:37:00 <karsten> okay, cool!
14:37:15 <karsten> moving on?
14:37:17 <irl> yep
14:37:21 <karsten> * ExoneraTor review #27356 (karsten)
14:37:31 <karsten> any plans for that?
14:37:38 <karsten> (it's not a quick review.)
14:38:08 <irl> no current plans
14:38:21 <karsten> okay. if you can, maybe try to squeeze that in somewhere.
14:38:38 <hiro> irl: I can create the repo, and then configure the hooks on torgit side
14:38:49 <karsten> hi hiro!
14:39:05 <irl> karsten: is it more or less important than #27390?
14:39:26 <karsten> #27390 is a much easier change, so could easily come first.
14:40:04 <karsten> (only thinking from a shortest-job-first perspective here..)
14:40:18 <irl> ok cool
14:40:37 <irl> hiro: i will think out a mail and send this to you because it would be for multiple repos
14:40:55 <irl> (also i will cc metrics-team@ so others can follow along)
14:41:21 <esaul> hi
14:41:30 <karsten> alright, I guess that also answers the #27390 topic then.
14:41:32 <karsten> esaul: hi!
14:41:52 <irl> karsten: yes (:
14:41:56 <karsten> 14:30:46 <+karsten> https://storm.torproject.org/shared/5h1Goax5eNusxjXJ_Ty5Wl7hFR1uqCReUiN8xdlBG8T <- agenda pad
14:42:00 <karsten> esaul: ^^
14:42:14 <esaul> thanks!
14:42:24 <karsten> okay, moving on to the next item.
14:42:30 <karsten> * Roadmap update (karsten)
14:42:33 <irl> https://trac.torproject.org/projects/tor/wiki/org/teams/MetricsTeam#RoadmapfromOctober2017toSeptember2018
14:43:35 <karsten> there, I just copied the roadmap to the pad.
14:44:00 <karsten> shall we go through the list?
14:44:03 <hiro> irl: sure no problems
14:44:31 <irl> ok
14:44:36 <karsten> so, 1.
14:45:01 <irl> did we undefer this one?
14:45:05 <karsten> we said we'd defer that, but I made more progress anyway after we decided we're not going to continue working on sponsor 13 deliverable 1.
14:45:14 <karsten> well, not continue until end of september.
14:45:18 <irl> right
14:45:47 <karsten> so, we did clean up some code and get rid of legacy code using pre-java-8 constructs.
14:46:07 <karsten> what we did not do is replace dates/times with java 8 types.
14:46:19 <karsten> nor did we replace directory traversal code, for example.
14:46:25 <karsten> I'd say we finished 50% here.
14:46:33 <irl> thinking about it in terms of time required, we did the easy bits but not the big hard bits
14:46:34 <karsten> and we won't do much mure.
14:46:38 <irl> so i would agree
14:46:41 <karsten> true.
14:47:18 <karsten> 2 was already at 100%.
14:47:21 <karsten> what about 3?
14:47:30 <karsten> I guess nothing has changed yet.
14:47:41 <irl> no progress that i can remember making
14:47:43 <karsten> but is this something we could still do in the next 2.5 weeks?
14:47:49 <irl> probably yes
14:47:50 <karsten> or part of?
14:48:10 <karsten> ok.
14:48:21 <karsten> no progress on 4, right?
14:48:52 <irl> no progress
14:48:56 <karsten> but should we try to move 4 forward?
14:49:05 <karsten> maybe in mexico? :)
14:49:14 <karsten> except that I won't have keys with me.
14:49:17 <karsten> hmmm.
14:49:19 <irl> i'm slowly getting more familiar with the services through osmosis
14:49:23 <irl> so i guess maybe 1%
14:49:26 <karsten> hehe
14:49:49 <karsten> so, okay, no progress is what counts for today.
14:50:01 <karsten> 5 was already done.
14:50:04 <karsten> what about 6?
14:50:26 <irl> 6 has been generating events in huginn for a while and i should sanity check them and either turn it on for real or go back and fix it
14:50:41 <karsten> is 80% still accurate?
14:50:43 <irl> no progress this week, maybe 100% by next week
14:50:46 <karsten> ok.
14:51:36 <karsten> 9 is in much better shape now.
14:51:53 <karsten> it's rebased to master, tested, migration is complete.
14:52:00 <karsten> what remains is review and deployment.
14:52:17 <karsten> if 75% sounds too much, we can pick a different number between 50 and 75.
14:52:28 <karsten> but the hard parts are out of the way, I believe.
14:52:30 <irl> 75% sounds ok
14:53:07 <karsten> 100% for 19 sounds good.
14:53:21 <karsten> what about 10?
14:53:27 <irl> no progress
14:53:29 <karsten> ok.
14:53:40 <karsten> and 12?
14:53:44 <irl> no progress
14:53:47 <karsten> ok.
14:54:11 <karsten> 13.
14:54:14 <karsten> -> no progress, but could be 100% next week
14:54:22 <irl> is there something to review for that?
14:54:35 <karsten> no. 2 things remain there:
14:54:49 <karsten> 1) link the new pages from more places on tor metrics.
14:55:06 <karsten> 2) include a survey for folks reading the reproducible metrics page to give us feedback.
14:55:26 <karsten> where 2) was a suggestion made by isabela, which I think makes sense here.
14:55:26 <irl> ok
14:55:32 <irl> sounds good
14:55:54 <karsten> ah, and creating the survey would bring us to 100%, not responding to feedback coming out of that.
14:56:04 <irl> ok
14:56:09 <karsten> we'd run it for a few weeks or even months and create new tickets for feedback.
14:56:43 <karsten> 14. hmm, no progress, would rather want to defer.
14:56:53 <karsten> or even discuss whether it's such a good idea after all.
14:57:06 <karsten> which we could do it in mexico.
14:57:35 <karsten> also, no progress on 15.
14:58:02 <karsten> 16 and 17 can advance because of time advancing. not sure if this was smart to add to the roadmap, but anyway.
14:58:39 <karsten> and no progress on 18.
14:58:51 <irl> for the next roadmap i think we can just reduce the amount of time we have for tasks and have the supporting core tor/tbb be implied
14:58:52 <karsten> I think that's all.
14:58:57 <karsten> right.
14:59:14 <irl> ok, and we've run out of roadmap (:
14:59:19 <karsten> yep!
14:59:36 <karsten> I'll update the wiki page after the meeting.
14:59:39 <karsten> next one:
14:59:40 <karsten> * Monthly report for August (karsten)
14:59:44 <esaul> Is there anything I can do to help?
14:59:53 <irl> the draft that you sent looked good
14:59:57 <irl> i did reply, but maybe it was lost
15:00:20 <karsten> esaul: let's add that as a topic to the agenda.
15:01:04 <karsten> irl: was briefly confused which draft you meant (after reading esaul's message), but yes, your reply arrived.
15:01:15 <karsten> so, okay, I'll send it out after the meeting, too.
15:01:18 <irl> awesome
15:01:34 <karsten> * Onionoo release (karsten)
15:01:44 <karsten> that's already on the ticket.
15:01:49 <irl> #27442
15:01:53 <karsten> so, we're planning to release/deploy today.
15:01:58 <irl> yes
15:02:24 <karsten> okay, great!
15:02:29 <karsten> * Volunteering
15:02:33 <karsten> esaul: hello :)
15:02:39 <irl> esaul: hi!
15:02:54 <esaul> hey whats up you guys, I have been waiting for this meeting for a while
15:03:03 <karsten> heh, sorry for the delay.
15:03:11 <irl> heh, sorry i've been travelling and making meetings awkward to arrange
15:03:12 <karsten> we usually have this meeting once per week.
15:03:29 <karsten> we can even prove that with help of meetbot!
15:03:29 <esaul> Yeah its ok, i understand
15:03:54 <esaul> hahaha yeah, I've been paying attention to those archived chats
15:04:16 <karsten> so, I responded to that thread on tor-dev with some suggestions.
15:04:29 <karsten> was there anything in that thread you'd be interested in?
15:04:42 <karsten> or would you like to find something else?
15:05:17 * karsten finds the link.
15:05:21 <esaul> The python package seemed interesting, I just hope learning how the codebase isnt too much to understand it
15:06:03 <karsten> https://lists.torproject.org/pipermail/tor-dev/2018-August/013404.html
15:06:15 <esaul> Honestly the easier the task teh better, since this is my first contribution
15:06:29 <karsten> that makes sense. which python package did you mean?
15:06:51 <irl> was that the stem ticket?
15:07:09 <karsten> ah, we discussed that at one of the meetings, right?
15:07:12 <karsten> briefly.
15:07:15 <irl> #17979
15:07:39 <irl> there is not a lot of context on that ticket, but it is a python thing
15:07:44 <karsten> (I gotta run for 5 mins.)
15:08:17 <esaul> yeah sorry about that it was that one
15:09:00 <esaul> I actually wouldnt mind , reproducing your data
15:09:42 <irl> it would be cool if you could look at the reproducible metrics and actually reproduce them (:
15:09:56 <irl> i think that would be the best way to find bugs in our descriptions
15:10:28 <esaul> Would you day there would be a huge learning curve to reproduce the data?
15:11:08 <irl> not a huge learning curve
15:11:26 <esaul> Alright I'll take on the task
15:12:19 <irl> i guess where we say "Obtain bridge network statuses from CollecTor." there you might want karsten to give you the magic wget/curl command that does that
15:12:51 <esaul> gotcha
15:12:54 <irl> python-stem can parse the descriptors but can't download them automatically
15:13:01 <karsten> (back, reading backlog)
15:13:17 <irl> it might be fun to do a jupyter notebook for each metric, as an idea
15:13:41 <esaul> thats what I had in mind lol
15:13:45 <irl> awesome (:
15:14:05 <esaul> I was between jupyter and rmarkdown haha
15:14:07 <karsten> sounds like you made a good plan!
15:14:22 <esaul> haha
15:14:28 <irl> i've used jupyter before, but i've also been tempted by rmarkdown
15:14:44 <esaul> ohhhhhh
15:14:54 <esaul> I am down for either one
15:15:00 <irl> do both?
15:15:02 <irl> (:
15:15:14 <irl> we don't have an R parser for the descriptors though
15:15:24 <irl> so that would be an argument in favour of jupyter
15:15:41 <esaul> Jupyter it is then it seems
15:15:52 * irl looks forward to seeing the notebooks
15:16:02 * karsten is curious, too!
15:16:09 <esaul> do you guys have a due date in mind?
15:16:38 <karsten> we're soon going to link the page from more places on the website.
15:17:05 <karsten> so, if there are easy ways to make it easier to comprehend, the sooner the better.
15:17:25 <karsten> like, we might add those links next week.
15:17:41 <esaul> gotcha
15:17:41 <karsten> of course, it will take longer to reproduce everything on that page.
15:18:08 <karsten> great!
15:18:35 <esaul> Anyhting else ? Is there a best way to contact you guys?
15:19:00 <irl> the metrics-team@ mailing list is a good place for async contact
15:19:30 <karsten> and no worries about putting too much noise on it. it doesn't have that many subscribers, and it's low volume.
15:19:42 <karsten> as long as a discussion is related, it can go there.
15:19:52 <esaul> Great!
15:20:02 <karsten> cool. thanks!
15:20:27 <karsten> alright. anything else for today?
15:20:31 <karsten> time for next meeting?
15:20:53 <karsten> when is mexico city...
15:21:04 <irl> mexico is the end of the month
15:21:28 <karsten> I wonder, should we stick with mondays until mexico city?
15:21:41 <irl> we can do
15:21:44 <karsten> not sure how useful the 27th would be, and then we'd be down to 1 more meeting this month.
15:21:52 <karsten> when switching back to thursdays, that is.
15:22:00 <esaul> YOU guys are going to mexico!!!?1!?!
15:22:00 <karsten> mondays would be the 17th and the 24th.
15:22:10 <karsten> we are!
15:22:24 <esaul> Lucky hahah
15:22:44 <karsten> the next tor meeting happens there.
15:22:49 <irl> 17th and 24th look good
15:22:54 <karsten> https://trac.torproject.org/projects/tor/wiki/org/meetings/2018MexicoCity
15:22:59 <esaul> im so jealous
15:23:39 <karsten> heh. if only there wouldn't be this jetlag...
15:23:49 <karsten> anyway, 17th and 24th then.
15:23:52 <esaul> Is there anyone I can talk to about employment at Tor?
15:24:23 <karsten> I think open positions are posted on the website somewhere.
15:24:26 <irl> https://www.torproject.org/about/jobs.html.en
15:24:27 <karsten> the tor website, that is.
15:24:29 <karsten> heh, there.
15:24:45 <esaul> Thanks!
15:25:08 <karsten> alright, let's end this meeting and proceed with the post-meeting!
15:25:24 <irl> ok
15:25:34 <karsten> thanks, everyone! :)
15:25:40 <karsten> #endmeeting