14:00:03 <karsten> #startmeeting metrics team
14:00:03 <MeetBot> Meeting started Thu Jun 30 14:00:03 2016 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:03 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:00:08 <oma> hi
14:00:15 <karsten> hi! who's here for the meeting?
14:00:28 <Letty> hi oma
14:00:28 <karsten> I've seen anathema_db, Letty, and now oma. hi!
14:00:52 <Letty> hello!
14:01:03 <karsten> https://pad.riseup.net/p/zUNzEIFRq5S4 <- agenda pad, not yet populated for today's meeting
14:01:19 <karsten> please add topics!
14:02:49 <anathema_db> k
14:05:17 <karsten> okay, 5 minutes into the meeting, let's start with what he have on the agenda.
14:05:21 <karsten> * MOSS award (karsten)
14:05:36 <karsten> https://blog.torproject.org/blog/tors-innovative-metrics-program-receives-award-mozilla
14:05:59 <karsten> just in case you haven't seen it on the blog or the team mailing list. :)
14:06:05 <karsten> or twitter
14:06:12 <Letty> yay!
14:06:31 <karsten> the status is that we're waiting on them to get the paperwork done, and then we can make some real plans.
14:06:55 <karsten> to be clear, this whole award thing went really, really fast. we applied on may 31.
14:07:19 <karsten> and it's the first round of MOSS funding other projects than mozilla-internal ones, I think.
14:07:38 <karsten> so it could take a few weeks for the paperwork to be written and signed and all that.
14:08:04 <karsten> I'll keep you posted.
14:08:09 <anathema_db> I'm kind of 'intruder' here, but I'd like to congrats with everyone involved in the project and the award. Well done!
14:08:26 <karsten> thanks! :)
14:08:49 <karsten> next time you'll be one of the insiders who receive congrats. ;)
14:09:15 <karsten> okay, are there any questions about the MOSS award that we should discuss here?
14:09:50 <Letty> not me, we talked already about that
14:10:09 <anathema_db> as you said, we should wait for make some real plans, right ?
14:10:31 <karsten> yes, give it a few weeks, I'd say.
14:10:33 <anathema_db> just wondering if there was/is already a roadmap, a 'practical' one
14:10:36 <anathema_db> cool
14:10:53 <karsten> there's one sketched on paper, yes.
14:11:00 <anathema_db> ahah good old paper
14:11:18 <karsten> the 10000-feet version is that we're spending roughly 2 months on each of the 6 deliverables. (surprise?!)
14:11:42 <karsten> but in a sense that we stop doing more because we're in the flow and instead head on to the next deliverable.
14:12:08 <karsten> but I guess I'll have a much better plan in a week or two, ideally after that paperwork is done.
14:12:25 <karsten> okay, moving on:
14:12:28 <karsten> * Berlin meeting in August/September (karsten)
14:12:50 <karsten> we're planning to have a meeting with metrics people and other tor people.
14:13:03 <karsten> not the big dev meeting in seattle, but a small gathering in berlin.
14:13:16 <karsten> here's the doodle if you're considering to join us: http://doodle.com/poll/tziw2ikwvgw2w8tf
14:13:47 <karsten> and there's an email on metrics-team@ with a little more context.
14:14:06 <karsten> https://lists.torproject.org/pipermail/metrics-team/2016-June/000150.html <- this one
14:14:56 <karsten> any questions about this meeting? it's still very early in the planning phase, so we might not find good answers yet. but the earlier we know about the questions, the better.
14:15:29 <anathema_db> is there an idea on the place to rent for the meeting? like hotel, conference room, $random ?
14:15:53 <karsten> the meeting will very likely be in the onion space office.
14:16:15 <oma> the planning is 3 days for metrics stuff?
14:16:17 <anathema_db> ?
14:16:23 <anathema_db> "onion space office" ?
14:16:53 <karsten> okay, there's not much on the internets about the onion space office. :)
14:17:10 <karsten> so, it's an office with 1 large room and 2 small rooms, run by tor people and friends.
14:17:21 <karsten> in wedding.
14:17:47 <karsten> but, not 100% certain about that yet. it may depend on the number of people.
14:18:08 <karsten> oma: not 3 days for metrics stuff, no.
14:18:27 <karsten> and metrics folks are not expected to stay for 3 days, unless they want to.
14:18:44 <karsten> it's also a meeting of tor folks who can't make it to seattle.
14:19:14 <anathema_db> is there already an agenda or something for the meeting?
14:19:24 <oma> somehow related: who has already made up it’s mind about seattle?
14:19:31 <karsten> and it might be that it will be only 2 days, not 3. but it's easier to plan for 3 days and shrink to 2 than the other way round.
14:19:51 <karsten> anathema_db: no agenda yet. but we can collect ideas now.
14:20:02 <Letty> i'm not in seattle
14:20:41 <karsten> oma: I think the planning of the seattle meeting is still in progress, too.
14:20:53 <anathema_db> karsten: sure. I think for metrics-related stuff we can wait for the roadmap related to MOSS award. We can use the meeting to discuss about critical/main points
14:20:58 <anathema_db> and maybe coding hard.
14:21:36 <oma> i’m about 75% decided that i’ll go to seattle
14:22:27 <karsten> anathema_db: right, the roadmap will help with that. and we're only 2 months into the 12 month project when we meet in berlin, so we can still make the roadmap better then.
14:22:43 <anathema_db> agree
14:23:16 <karsten> so, I think the next step is that more people participate in the doodle and GeKo closes that, say, next week.
14:23:56 <oma> when do the 12 months of the MOSS project begin? with the signage of the paperwork?
14:24:00 <karsten> and let's just come up with possible agenda items while we think of them and later add them to a wiki page.
14:24:07 <karsten> oma: if only I knew..
14:24:24 <karsten> oma: my hope is on july 1, because I want to include current improvements in the final report,
14:24:34 <karsten> but I can't say for certain.
14:24:42 <oma> kasrten: jsut hold them back :-)
14:25:00 <karsten> you mean like the past 5 years? ;D
14:25:02 <oma> oups … karsten: just hold them back
14:25:16 <karsten> nah, I hope to hear back from them in the next few days.
14:25:18 <oma> just another few days … ;-)
14:25:26 <karsten> okay, moving on?
14:25:37 <karsten> * Disagreement integration & tool for code styling? (letty)
14:25:54 <karsten> how's that going, and what tool are you looking for?
14:26:08 <Letty> its more like questions.. karsten you told me about a tool for code formatting. with the google stye guides included
14:26:14 <karsten> yes!
14:26:25 <karsten> we're using that for collector now. let me find a link for you.
14:26:43 <Letty> i would like to have it :D before i send you my messy code
14:27:17 <karsten> https://gitweb.torproject.org/collector.git/commit/?id=1609d64dbfa856943947fed5fe7c5b42ecc9446f
14:27:48 <anathema_db> what I'm looking at?
14:27:54 <karsten> checkstyle.
14:28:05 <oma> interesting
14:28:25 <karsten> Letty: maybe it's easier to just use eclipse to auto-format the code.
14:28:32 <karsten> or whichever IDE you're using.
14:28:45 <Letty> ahh there is a eclipse plugin
14:29:26 <anathema_db> sorry, I think I'm lost.
14:29:35 <karsten> https://trac.torproject.org/projects/tor/wiki/org/teams/MetricsTeam/MetricsJavaStyleGuide <- also relevant
14:29:57 <anathema_db> are we talking about using a formatting guide for our code?
14:29:58 <oma> anathema_db: Checkstyle configuration that checks the Google coding conventions from Google Java Style
14:29:59 <oma> +    that can be found at https://google.github.io/styleguide/javaguide.html.
14:30:00 <oma> +
14:30:01 <karsten> anathema_db: checkstyle checks your code for formatting problems. we're using google's style guide.
14:30:01 <oma> +    Checkstyle is very configurable. Be sure to read the documentation at
14:30:02 <oma> +    http://checkstyle.sf.net (or in your downloaded distribution).
14:30:15 <Letty> thx!
14:30:23 <anathema_db> aah ok now it makes more sense, thanks
14:30:38 <karsten> sure. still work in progress, but ideally we'll have most code checked by checkstyle in the future.
14:30:50 <Letty> karsten: another question is, do you have time for code review? next week or so?
14:31:14 <karsten> sure, just send the code, and I'll squeeze it in somewhere.
14:31:25 <Letty> ok great
14:31:26 <anathema_db> that's just ffor java right?
14:31:34 <Letty> next topic
14:31:38 <anathema_db> what if python or other?
14:31:39 <Letty> yes
14:32:09 <karsten> anathema_db: good question. I think there are similar guides.
14:32:23 <karsten> https://google.github.io/styleguide/pyguide.html maybe?
14:32:33 <karsten> and maybe there are similar tools to check style in python?
14:32:44 <oma> ah, next topic is me.
14:32:50 <karsten> hang on
14:32:53 <anathema_db> I use PEP8 as coding style, automatically checked by Pycharm
14:33:36 <karsten> anathema_db: we don't have much python code in metrics-related tools, which is why we didn't focus on that yet, but if you'd want to write something about that, we'd gladly add that to
14:33:39 <karsten> https://trac.torproject.org/projects/tor/wiki/org/teams/MetricsTeam/Documentation
14:33:49 <karsten> see "(add more here, when need arises) "
14:34:33 <karsten> first sets the rules, second needs to fight them.
14:34:34 <Letty> anathema_db: is that an ide? i use for javascript my build in function from webstorm. thats perfect. better than no styling
14:36:15 <anathema_db> Letty: yes it's a java-based IDE for python
14:36:34 <anathema_db> but you can use it for different file formats
14:36:34 <karsten> okay, shall we move on?
14:36:41 <anathema_db> yep
14:36:48 <karsten> cool!
14:36:51 <karsten> * short analytics server update (oma)
14:37:17 <oma> analytics server update: converTor tool is nearly finished (need to incorporate iwakehs code review, but wait for him to finish the code review), second round of conversion will start after that as Karsten has already updated the descriptors. The next big task is doing some useful aggregations. Setting up an environment with Spark and an IDE. Did taht too. Now need to learn a littel Scala
14:37:58 <karsten> there will be new bridge descriptor tarballs soon.
14:38:05 <karsten> well, next week.
14:38:17 <karsten> it's still chewing on them.
14:38:37 <karsten> and there will be a metrics-lib release as soon as I know whether that counts as improvement for MOSS.
14:38:38 <oma> those were still missing? okay. just ping me when they arrive, please
14:38:46 <karsten> well,
14:39:05 <karsten> we changed the sanitization algorithm.
14:39:11 <karsten> let me find the ticket.
14:39:32 <karsten> https://trac.torproject.org/projects/tor/ticket/19317
14:39:47 <karsten> where changed means extended by yet one more thing we should sanitize.
14:40:02 <karsten> (tcp ports)
14:40:18 <oma> okay, right now I have enough data for testing/prototyping. No need to hurry with releases etc
14:40:24 <karsten> I'm also considering splitting up tarballs into bridge-statuses, bridge-server-descriptors, and bridge-extra-infos.
14:40:29 <karsten> ok.
14:40:36 <oma> that would be good I think
14:41:01 <karsten> yes. and now I finally have an excuse for doing it.
14:41:10 <oma> right!
14:41:35 <karsten> okay, what else do you need?
14:42:23 <oma> oh, nothing for the nexrt few weeks. learning scala and prototyping aggregations will keep me busy.
14:42:35 <karsten> okay.
14:42:48 <oma> i hope to get to teh point where i can aggregate everything i used in visionion
14:42:57 <oma> next topic?
14:43:02 <karsten> sure:
14:43:03 <karsten> * testing converTor (oma)
14:43:25 <oma> iwakeh brought it up in his code review again
14:43:58 <oma> karsten, we talked about it briefly and I went away with the impression that it’s a major task
14:44:43 <oma> what i see is that i would have to build sample descriptors of everey type with every possible value
14:44:59 <oma> and then run the convertor over them
14:45:03 <oma> is that right?
14:45:20 <isabela> oi
14:45:43 <karsten> oma: basically, you'd have to provide test input and be able to check test output.
14:46:14 <oma> and providing the test input is where the problem lies, right?
14:46:14 <nickm> morning
14:46:16 <karsten> and you'd need a way to encapsulate the unit you're testing by making it as independent of the environment as possible (system time, file system, etc.).
14:46:29 <nickm> teor: thanks!
14:47:05 <karsten> hi isabela and nickm. we'll give back the channel soon. sorry! :)
14:47:52 <karsten> oma: depends. providing good test input can be hard. but even if you start with the test input that's easy to produce, and you raise test coverage from 0 to, say, 50%, you gained something.
14:48:34 <karsten> oma: and making your code more testable has advantages, too.
14:49:14 <karsten> oma: but I can also see that you'd first want to see if the output is useful at all before making sure it's computed 100% correctly.
14:49:15 <oma> karsten: and should I look for a test framework? JUnit? or has Java something build in?
14:49:20 <karsten> junit, yes.
14:49:52 <oma> something i should know about junit, anywhere special i should start?
14:50:29 <Letty> if you want to become familiar. just start a little projekt and test basic function (eg calculations)
14:50:32 <karsten> there are plenty of junit books and tutorials available.
14:50:33 <oma> (sorry, special ;-) I meant "specific")
14:51:03 <oma> okay, so nothing specific. alright
14:51:27 <oma> <done/>
14:51:44 <karsten> ok. before me move on to the last item, which is going to be quick,
14:51:52 <karsten> anything else to discuss this week?
14:52:36 <karsten> ok.
14:52:37 <karsten> * Meeting schedule (karsten)
14:52:49 <karsten> should we change the schedule to weekly?
14:53:05 <karsten> One change that iwakeh and I were considering is to have a weekly
14:53:06 <karsten> meeting instead of one happening every other week.  That would make it
14:53:06 <karsten> less bad to miss a meeting every now and then, and we wouldn't have to
14:53:06 <karsten> reschedule meetings on possibly short notice.  Happy to discuss more
14:53:06 <karsten> tomorrow or at the meeting after that.
14:53:22 <karsten> (from my meeting announcement)
14:53:32 <karsten> what do you think?
14:54:00 <anathema_db> it would be not a bad idea
14:54:05 <anathema_db> the meeting would be quicker
14:54:10 <karsten> hehe
14:54:16 <anathema_db> and as you said, missing one would be not a big deal
14:54:17 <oma> if it’s weekly i’ll miss more meetings…
14:54:22 <karsten> right.
14:54:31 <karsten> there would be logs.
14:54:32 <anathema_db> oma: why if I may ask ?
14:54:49 <Letty> i agree with anathema
14:55:00 <oma> anathema_db: because of other commitments
14:55:18 <anathema_db> but you're not going to miss much
14:55:33 <anathema_db> if every 2 weeks we are able to do 1h meeting
14:55:42 <anathema_db> doing every week it would be…30mins ? :)
14:55:59 <anathema_db> or maybe we can find more stuff to talk about
14:56:00 <karsten> well, I don't think we can go to 30 minutes.
14:56:09 <anathema_db> karsten: was just an example
14:56:11 <karsten> we can try, of course.
14:56:38 <karsten> I mean, we could try to set the time to 14:00--14:45, for example.
14:56:49 <karsten> mostly as a sign that we don't want to spend twice as much time meeting.
14:56:50 <oma> i’m just saying. and it doesn’t feel like the bi-weekly meetings arr overly cramped with stuff we have to talk about. also a 30-minutes-meeting means i have to make sure i’m not AFK etc
14:56:55 <anathema_db> to say: if now we can schedule a meeting every 2 weeks due to workload, scheduling 1 every week it would be more like a a quick update
14:57:50 <karsten> how about we try weekly meetings in july, with a limit of 45 (or 30?) minutes?
14:57:57 <karsten> and then evaluate at the end of july?
14:57:59 <oma> I don’t want to be the nay sayer :-)  let’s try it?
14:58:02 <anathema_db> sounds good
14:58:18 <Letty> i think weekly is perfect for people who work a lot on metrics stuff. you dont have to wayit two weeks or just writing emails with the other person directly. maybe you will have also some good suggestions form other people around
14:58:28 <anathema_db> 45 minutes that can be shrank to 30
14:58:35 <karsten> sold!
14:58:47 <anathema_db> sorted!
14:59:02 <anathema_db> (as they say in Ireland)
14:59:05 <karsten> okay, great, thanks everyone for attending! talk to you in 1 week from now.
14:59:15 <anathema_db> cool! thanks everyone
14:59:21 <karsten> sorted it is!
14:59:27 <anathema_db> :D
14:59:42 <karsten> #endmeeting