13:59:46 <karsten> #startmeeting Measurement Team meeting
13:59:46 <MeetBot> Meeting started Wed Aug 19 13:59:46 2015 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:59:46 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
13:59:53 <karsten> who's here for the meeting?
13:59:55 * phw 
14:00:09 <karsten> hello, phw.
14:00:40 <arma4> i'm accidentally here
14:00:50 <karsten> while we wait, here's the 1-1-1 task exchange pad of the week: https://etherpad.mozilla.org/czd5Jku3DR
14:01:14 <karsten> hi arma4.
14:01:57 <qbi> I'm a lil bit here, but mostly in another meeting. ;)
14:02:15 <karsten> suggested agenda for today: do another round of the 1-1-1 task exchange game and talk about whatever measurement-related things people want to talk about.
14:02:32 <karsten> hi qbi. just focus on whichever meeting is more fun. ;)
14:03:07 <karsten> do others have suggestions for the agenda?
14:03:37 <karsten> like, should start meetings in the future with update rounds?
14:03:42 <arma4> your task exchange game looks fun
14:03:44 <karsten> should we*
14:04:02 <karsten> it is. think about something you want to get done and add it.
14:06:17 * karsten puts "should we do update rounds" on the agenda as a discussion point for today.
14:06:20 <arma4> am i being tricked into receiving a task, by doing so? ;)
14:06:39 <karsten> nah. :)
14:07:31 <arma4> what's the story on metrics reporting client count by PT type? is the current issue that bridges don't even collect this info, so of course metrics cannot graph it for us?
14:07:53 <arma4> karsten: also, i had a good discussion with tariq last week, where he is running privex on some exit relays, but he believes he is doing it safely
14:07:59 <karsten> yes, and that it's unclear how bridges would collect that info safely.
14:08:05 <arma4> i pointed him at all the guidelines for safe and ethical stats research
14:08:16 <arma4> and he's going to jump in and try to clean those guidelines up
14:08:33 <karsten> sounds great!
14:08:40 <arma4> it pretty clearly lines up with sponsor q also
14:08:58 <arma4> if we know anybody who wants to help make those guidelines complete and also easy to read, that'd be grand
14:08:58 <karsten> I have a similar to-do list item here, but it didn't make it to the top yet.
14:09:03 <karsten> (^ guidelines)
14:09:14 <arma4> one of them is "don't collect anything you don't want to publish". that's the one you put in.
14:09:48 <arma4> there's another, which is "add noise so you get the long-term answer to the question you wanted to ask, without giving useful details about finer granularity"
14:10:25 <arma4> there was a third big one. i wonder what it was.
14:11:00 <karsten> I wonder what we wrote in that case study paper together with steven. years ago.
14:11:28 <karsten> so, yes, step 1 would be to collect all these ideas/guidelines.
14:12:07 <karsten> isabela and I talked about this in boston and how we could even turn it into a blog post and show the world that we care about this stuff.
14:12:41 <isabela> we should do it!
14:12:44 <isabela> sorry :)
14:12:54 <karsten> hi isabela. :)
14:13:18 <karsten> yes, we should. I'll think more about this and maybe start with step 1 this week.
14:13:47 <karsten> okay, should we exchange some tasks?
14:13:52 <karsten> https://etherpad.mozilla.org/czd5Jku3DR
14:14:13 <karsten> I already added two last week, but I'll add another one to give yet one more example.
14:14:20 <karsten> Be the first person to review the concept of using data-aggregating modules to provide data for the Metrics website; the goal here is to make it easier for others to contribute new data aggregations to the Metrics website and at the same time feasible for us to run them.
14:14:24 <karsten> https://people.torproject.org/~karsten/volatile/metrics-modules.md
14:14:44 <karsten> (that's just a quick draft that I'll clean up more in the next 24 hours.)
14:15:40 <karsten> anyone want to add something?
14:17:13 <karsten> arma4: I'll add your guidelines idea as another task and will grab it, for another example of a good task.
14:19:48 <karsten> phw: btw, how did that work out with sean's task?
14:20:18 <phw> yesterday, i sent sean an email with a bunch of ideas.
14:20:20 <karsten> qbi: thanks again for helping resolve those two ExoneraTor tickets!
14:20:33 <karsten> phw: ah, perfect.
14:21:00 <qbi> np.
14:21:08 <karsten> speaking of, I asked sean to add a badge for whether a relay operator should have a t-shirt, and he liked the idea.
14:22:46 <karsten> phw, qbi, others: want to add another task?
14:23:46 <qbi> no
14:26:19 <karsten> phw: while you're around, mind if we quickly discuss how to move #15178 forward?
14:27:13 <karsten> phw: as in, would it be a problem if I changed Onionoo like I suggested on that ticket?
14:28:10 <karsten> one possible downside is that if onionoo stops updating, and I don't fix it, atlas will present its outdated data as if that was the current data.
14:28:49 <arma4> karsten: somewhere in trac is a ticket for writing up guidelines on research.tp.o.
14:29:04 <karsten> arma4: looking..
14:29:20 <arma4> i filed it, or maybe phw did, during the fuss with andrei
14:32:18 <karsten> hmm, I found a riseup pad created by phw, but that's empty now.
14:32:30 <karsten> (found in my inbox)
14:33:35 <phw> back again
14:33:49 <phw> sure, let's discuss.
14:34:46 <karsten> #16532 !
14:35:07 <karsten> phw: ok.
14:35:27 <karsten> phw: what do you think, would that make atlas users sad?
14:35:43 <karsten> (I shall also ask isis what she thinks about this change, but I need to start somewhere.)
14:36:06 <karsten> (this is about #15178)
14:37:42 <phw> sounds good to me.  but i probably won't manage to work on that in the near future.
14:38:35 <karsten> well, would you rather want me to wait with making this change until you have the atlas change in?
14:39:20 <karsten> the ticket has been sitting there for 5 months, so we might be able to wait until your distant future starts.
14:40:01 <phw> also fine with me.
14:40:11 <karsten> because right now, atlas users do get feedback when data is outdated: server error. they can then walk over to atlas.thecthulhu.com.
14:40:59 <karsten> okay, let's wait then. I'll ask isis anyway to hack on the equivalent in globe.
14:41:32 <karsten> should we close this round of 1-1-1's and leave tasks in place for next week when we might have more people?
14:42:21 <phw> i have nothing to add.
14:42:30 <karsten> okay.
14:42:48 <karsten> "update rounds" <- do we want to do that in future meetings, or are we okay without doing them?
14:44:43 <karsten> otherwise, do we have anything else to discuss today?
14:45:06 <phw> whoever wants to give an update could do so.  i'm not sure if our meetings have to be so strictly defined.
14:45:17 <phw> especially given that we are usually 2-3 people.
14:45:50 <karsten> yes, I agree. works for me.
14:47:07 <karsten> cool, I'd say let's end this meeting early, unless somebody comes up with a topic in the next minute or so.
14:48:35 <karsten> #endmeeting