14:32:28 <karsten> #startmeeting metrics team
14:32:28 <MeetBot> Meeting started Thu Mar  1 14:32:28 2018 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:32:28 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:32:33 <karsten> hello!
14:32:45 <karsten> https://storm.torproject.org/shared/Oh4g0hNenh65QZNWRsIe5zxpB3e0axASeSgo5hKOp2A <- agenda pad
14:33:58 <karsten> irl: want to add something to the agenda pad, too?
14:34:19 <karsten> it just looks like a lot, because it contains the roadmap update draft and monthly report draft.
14:34:25 <karsten> but it's just four topics.
14:35:01 <irl> added a thing
14:35:04 <karsten> :)
14:35:08 <iwakeh> five
14:35:16 <karsten> okay, shall we start?
14:35:22 <iwakeh> yep.
14:35:28 <irl> ok
14:35:32 <karsten> * goals for further rs integration into metrics-web, see https://trac.torproject.org/projects/tor/ticket/25393 (iwakeh)
14:35:55 <iwakeh> While looking at #22423 I noticed
14:36:31 <iwakeh> the duplications and other tasks left from the integration.
14:37:18 <iwakeh> We should discuss and decide
14:37:23 <iwakeh> Should RelaySearch be developed as a part of metrics.tp.o that can also be served stand-alone, which is the current situation or just as a part of metrics.tp.o?
14:37:33 <iwakeh> at some point.
14:37:37 <karsten> that's a good question. it's also relevant for exonerator.
14:37:45 <iwakeh> true.
14:37:59 <karsten> it ties into maintaining more things.
14:38:13 <karsten> standalone + integrated = 2 things.
14:38:32 <iwakeh> and
14:38:33 <karsten> same as pre-aggregated CSV files + per-graph CSV files.
14:38:34 <irl> a quick google for tor atlas doesn't find any other hosted instances
14:38:36 <iwakeh> interdependencies
14:38:48 <irl> nor does tor relay search
14:38:57 <iwakeh> Well, these could always use metrics-web
14:39:09 <iwakeh> and remove the stuff they don't want to host.
14:39:13 <iwakeh> :)
14:39:25 <irl> i think in the longer term i would like to stop treating it as a submodule and integrate it fully
14:39:40 <iwakeh> fine, that makes
14:39:45 <irl> #12522 is one motivation for this
14:39:49 <iwakeh> #25392 also easier.
14:40:28 <irl> currently there are javascript libraries that need to be moved from js/libs/* and integrated properly, and fonts too as you've seen
14:40:44 <iwakeh> yes.
14:40:58 <iwakeh> I was going to start with the metrics-web side
14:41:09 <irl> i wonder if a good time to do this would be when i do the change over to bootstrap 4.0.0 and the new tor styleguide
14:41:17 <iwakeh> of 25392
14:41:19 <irl> as that would mean changing a number of libraries anyway
14:41:39 <irl> i have no objection to things like fonts being fixed straight away
14:41:44 <iwakeh> ah, ok.
14:41:56 <irl> if you can diff files and they're identical then i would have no objection to removing those duplicates either
14:42:05 <irl> e.g. for js libs
14:42:24 <iwakeh> yes, I could adjust the project layout and Ant tasks and then
14:42:37 <irl> that sounds fine to me, and i can review those changes
14:42:42 <iwakeh> start to identify the issues like libs fonts etc. which
14:42:51 <iwakeh> we could figure out together
14:42:56 <irl> yep
14:43:16 <iwakeh> great
14:43:19 <irl> (:
14:43:19 <karsten> want to add some notes to the pad?
14:43:39 <iwakeh> basically, we agree on tight integration
14:43:46 <iwakeh> and no extra deployment?
14:43:52 <irl> yes
14:44:00 <iwakeh> adding ...
14:44:00 <karsten> it certainly makes things easier.
14:44:12 <karsten> what about exonerator?
14:44:39 <irl> exonerator is perhaps more complicated because we know of external deployments
14:45:06 <irl> i think we would upset people if we broke them
14:45:16 <karsten> ok.
14:45:22 <karsten> good point.
14:45:28 <irl> or even, they would stop using them and then treat all ip addresses equally
14:45:34 <irl> which would be a scary prospect for exit operators
14:45:43 <karsten> ah, that's not exonerator.
14:46:05 <karsten> it's the database for looking up whether an address was used as exit in the past.
14:46:14 <iwakeh> exonerator might be easier to 'extract' as stand-alone from metrics-web.
14:46:16 <karsten> or, a relay.
14:46:34 <karsten> alright, let's consider relay search separately from exonerator.
14:46:37 <iwakeh> so, there are no external ExoneraTors?
14:46:46 <karsten> well, maybe there are.
14:47:28 <karsten> anyway, trying to stay on topic.
14:47:35 <iwakeh> next?
14:47:43 <karsten> yep.
14:47:51 <karsten> * next steps webstats [b/c of https://lists.torproject.org/pipermail/metrics-team/2018-February/000706.html ] (iwakeh)
14:48:06 <iwakeh> Sebastian's mail.
14:48:11 <karsten> oh, yes.
14:48:34 <iwakeh> should the process be speeded up?
14:48:34 <karsten> what can we do?
14:48:44 <karsten> we're blocking on sysadmins.
14:48:52 <iwakeh> CollecTor needs the files.
14:48:56 <iwakeh> true.
14:48:57 <karsten> right.
14:49:24 <karsten> but even if we speed up the process, we won't be faster than sunday.
14:49:42 <karsten> realistically.
14:49:56 <iwakeh> Well, March would be ok?
14:50:06 <karsten> yes. that sounds realistic.
14:51:02 <karsten> sounds good.
14:51:11 <karsten> not much we can do now.
14:51:27 <iwakeh> yes. we prepared all.
14:51:33 <karsten> yep. moving on?
14:51:39 <iwakeh> fine.
14:51:44 <karsten> * Why metrics are important and what we do to make them safe #23169 (irl)
14:51:55 <iwakeh> cool!
14:52:04 <karsten> ah, needs review!
14:52:14 <irl> i was hoping to have an onion service to point you at
14:52:22 <irl> but i lost power with all the snow everywhere and i'm behind
14:52:30 <karsten> oh.
14:52:46 <irl> this is not meant to happen in the 3rd largest city in the country, we're meant to have reliable power
14:52:55 <irl> but i have managed to get all the text and graphics in that i wanted
14:52:55 <karsten> will take a look at the branch.
14:53:14 <iwakeh> Hope you're sitting in a warm place now!
14:53:24 * karsten experienced 4 days without power in one of the largest cities in the world. so....
14:53:34 <irl> it would be good if both karsten and iwakeh could review it, as i guess this is what people will read about the metrics team and it would be good if everyone is happy with the text
14:53:45 <irl> heh, oh dear
14:53:52 <irl> yes, i'm warm now
14:53:54 <karsten> sounds good. the part about reviewing.
14:54:20 <karsten> good to hear that, too. :)
14:54:23 <iwakeh> good.  (there must be some power source as you're participating)
14:54:57 <irl> also i guess relevant, #23830
14:55:15 <irl> iwakeh: yes, this is fine now. i can do ant war and then java -jar path/to/war and it just works.
14:55:34 <iwakeh> great, thanks!
14:55:34 <karsten> very useful!
14:56:06 <irl> ok, well i look forward to the reviews (:
14:56:15 <iwakeh> me2
14:56:25 <irl> for the next week i will mostly be doing OONI things but I should be able to respond to comments at least
14:56:57 <karsten> moving on?
14:57:00 <irl> ok
14:57:05 <iwakeh> yes
14:57:08 <karsten> * Roadmap update (karsten)
14:57:14 <karsten> I made some suggestions on the pad.
14:57:22 <karsten> want to take a look?
14:57:58 <iwakeh> should #24217 be updated?
14:58:23 <karsten> ah, yes.
14:58:39 <karsten> I mean, there's not much new to share.
14:58:59 <karsten> but yes, I'll write something there.
14:59:21 <iwakeh> fine
14:59:50 <iwakeh> I'll further look into 'external research data'
15:00:00 <iwakeh> integration.
15:00:11 <karsten> sounds good.
15:00:32 <iwakeh> and start thinking about 'Document the Tor Metrics system/pipeline'
15:00:48 <karsten> those two might be good candidates for the rome meeting week.
15:00:57 <karsten> I won't be online much during that week.
15:01:04 <iwakeh> or did you start on that?
15:01:08 <karsten> nope.
15:01:38 <iwakeh> true, I'll time these for that week.
15:01:54 <karsten> speaking of weeks, can you prioritize things that need my feedback next week?
15:02:08 <iwakeh> sure
15:02:15 <karsten> okay.
15:02:25 <iwakeh> what is most urgent currently?
15:02:43 <iwakeh> (beside what was mentioned)
15:03:00 <karsten> hmm, I think open reviews in general.
15:03:17 <iwakeh> ok, just the 'high'+ tickets.
15:03:36 <karsten> probably, yes.
15:03:56 <karsten> if I find something that is urgent and not yet on high, I'll bump it to high.
15:04:00 <karsten> before next friday. ;)
15:04:14 <iwakeh> I think the percentages are ok.
15:04:21 <karsten> I was just about to ask.
15:04:25 <iwakeh> on the roadmap.
15:04:26 <karsten> irl: what do you think about percentages?
15:04:48 <irl> i don't think i've moved any of them this week
15:04:55 <karsten> month
15:05:13 <karsten> last time we updated them was early february.
15:05:31 <iwakeh> (approx three weeks ago)
15:05:41 <irl> i mean, the +15 on relay search -> tor metrics is probably the only one
15:05:44 <irl> but that's already there
15:05:51 <karsten> ok.
15:05:59 <irl> hmm
15:06:13 <irl> maybe though atlas is now quite well renamed to relay search
15:06:18 <irl> so we can +5 on renaming components
15:06:27 <karsten> sure.
15:06:33 <irl> i did open approximate 1 gazillion bugs with patches for other things than link to the old URL
15:06:44 <karsten> heh
15:07:22 <karsten> any other changes?
15:07:49 <irl> not from me
15:08:00 <iwakeh> I'm fine.
15:08:03 <karsten> okay!
15:08:18 <karsten> I'll update the wiki after the meeting.
15:08:19 <karsten> * Monthly report draft (karsten)
15:08:26 <karsten> more on that on the pad.
15:08:30 <karsten> what's missing?
15:08:37 <karsten> what can be described better?
15:08:40 <iwakeh> looks complete now.
15:09:27 <irl> looks complete to me
15:09:33 <karsten> cool!
15:09:39 <irl> just deleted the "part" from the Relay Search sentence, it didn't parse for me
15:09:46 <karsten> I'll leave it on the pad for now and send the report tomorrow.
15:09:54 <karsten> heh, sounds good.
15:10:36 <karsten> okay, I think that's all.
15:10:51 <iwakeh> yes.
15:10:57 <karsten> shall we have a short meeting next week?
15:11:37 <karsten> I mean, we can't make big plans next thursday, as both irl and I will be traveling soon after.
15:11:43 <karsten> but just to sync?
15:11:46 <iwakeh> ah, ok.
15:11:55 <irl> i have no strong opinion
15:12:00 <irl> but i would be available
15:12:08 <karsten> okay, let's do it. usual time.
15:12:10 <iwakeh> me neither, we could decide by mail?
15:12:28 <iwakeh> yes, or just usual time.
15:12:35 <karsten> ok.
15:13:00 <karsten> thanks! have a good rest of a winter day. bye, bye!
15:13:05 <irl> bye!
15:13:10 <iwakeh> thanks! bye, bye.
15:13:16 <karsten> #endmeeting