15:58:57 <karsten> #startmeeting metrics team
15:58:57 <MeetBot> Meeting started Thu Feb 21 15:58:57 2019 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:58:57 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:59:28 <karsten> let's start.
15:59:37 <karsten> Roadmap:  ​https://storm.torproject.org/shared/TsgBadLSfM8uh_aftPjxmR_rm0a8E-4UQ2HEa_l0DHv
15:59:56 <karsten> irl: gaba and I moved around some things last week.
16:00:22 <irl> ok
16:00:34 <karsten> and we found these meetings to be a good time for moving things and then not touching them until the week after.
16:00:41 <karsten> like a plan for the week.
16:00:46 <irl> sounds good
16:00:53 <gaba> we created a column with issues we didn't recognize
16:00:56 <gaba> to discuss
16:01:00 <gaba> in this meeting
16:01:06 <karsten> right, and that.
16:01:23 <karsten> should we start with in progress?
16:01:31 <gaba> yes
16:01:45 <karsten> ok. so the first is what I have been working on.
16:01:54 <karsten> I'd like to discuss that in more detail later today.
16:02:03 <karsten> for now it's still in progress.
16:02:16 <karsten> the next is irl's.
16:02:20 <irl> we estimated 3 points for yours karsten
16:02:24 <irl> how many points has it been so far?
16:02:39 <karsten> 3 points == 1.5 weeks, right?
16:02:47 <irl> 1-1.5 weeks ish
16:02:47 <gaba> 3 points are 3 days
16:02:49 <gaba> 1 week
16:03:24 <karsten> 6 points so far?
16:03:34 <irl> oh dear
16:03:35 <karsten> which is why we'll have to talk about doing more.
16:03:39 <irl> ok
16:03:40 <gaba> ok
16:04:11 <karsten> but let's talk about that in a bit.
16:04:18 <karsten> if that's okay?
16:04:22 <irl> yeah
16:04:37 <karsten> ok. so your item has the last remaining report in it.
16:04:43 <irl> yes
16:04:48 <karsten> the collector prototype already has a report.
16:04:59 <irl> it is the codebase overviews and comparison to ooni
16:05:03 <karsten> right.
16:05:21 <irl> i think this has about 2-3 points left on the overviews
16:05:32 <irl> did we have any more input from ooni on the comparison?
16:05:39 <karsten> I didn't hear anything.
16:05:48 <karsten> I could try again.
16:05:56 <irl> sounds good
16:05:58 <gaba> were you talking with arturo?
16:06:12 <karsten> I did send him email, but I think leonid responded.
16:06:19 <gaba> ok
16:06:24 <karsten> I'll just ask once more.
16:06:36 <gaba> yes
16:07:12 <karsten> so, how many points if they respond? how many if not?
16:07:28 <irl> if they respond probably 2 points, if not then 4-5
16:07:55 <karsten> so, should we assign something now?
16:08:05 <karsten> 8 points, in the hope to get it down to 5?
16:08:15 <irl> sounds good
16:08:20 <karsten> great.
16:08:33 <gaba> yes, we need to be pesimist when estimating
16:08:35 <irl> we will not have more than 5 points done before the next meeting so it doesn't affect planning
16:08:48 <karsten> ok.
16:09:15 <karsten> that concludes in progress items.
16:09:43 <irl> do you want to look at the "to discuss"?
16:09:47 <karsten> yep.
16:09:56 <karsten> what's the first about?
16:10:05 <irl> #28465 is that i need to write a proposal to remove that line from the consensus
16:10:20 <irl> this will either trigger the dirauths to start using it so we keep it, or we just remove it
16:10:32 <irl> i need to start a discussion but don't necessarily need to participate in it
16:11:06 <karsten> does this need to happen now? as in, is there a deadline of some sort, so that it makes sense to do it soon?
16:11:35 <irl> it was something uncovered during the collector rewrite work, that we have this thing we have code in core tor for that is not used ever
16:11:49 <irl> it is probably a 15 minute task that can happen whenever
16:12:13 <gaba> we can have it in the backlog for the next month? or is this something for march?
16:12:26 <karsten> if it's 15 minutes, does it really belong on the roadmap?
16:12:30 <gaba> right
16:12:37 <irl> probably doesn't need to be on the roadmap
16:12:43 <gaba> 15 min is very low, i do not believe you irl :P
16:12:46 <karsten> haha
16:13:00 <karsten> I mean, there's a ticket for it, it won't get lost.
16:13:05 <gaba> yes
16:13:13 <karsten> it'll be in good company...
16:13:20 <irl> i might just do it this evening to be rid of it
16:13:24 <gaba> ok
16:13:25 <karsten> heh, okay.
16:13:35 <gaba> moving it to in progress
16:13:43 <irl> ok
16:13:53 <karsten> the next one is yours, too.
16:14:07 <irl> this is filling out a template that antonela has given me
16:14:18 <irl> in the process that 21 point ticket gets broken down into lots of smaller tickets
16:14:33 <karsten> ah, #24422?
16:14:35 <irl> yeah
16:14:50 <irl> i will also share the google doc with you when i start working on this
16:14:53 <irl> but it won't be this week
16:14:57 <karsten> okay. should we move it to in progress or icebox or backlog?
16:15:07 <irl> i think backlog
16:15:09 <karsten> (what was the difference again?)
16:15:23 <karsten> backlog first, then icebox?
16:15:37 <gaba> backlog is sort out for priority and is what we do next
16:15:46 <gaba> icebox are issues that need to happen but no necessary next
16:15:50 <karsten> okay.
16:16:02 <karsten> sure, backlog sounds fine.
16:16:05 <gaba> there is 24422 in backlog already
16:16:05 <irl> antonela needs the prd before the end of the month
16:16:12 <karsten> oh.
16:16:35 <karsten> so, then backlog makes even more sense.
16:16:48 <antonela> hey o/ -- that is what i say to simply secure, if is not doable, i can talk with them to think about it the next month
16:16:53 <gaba> moving it to the top of backlog then
16:16:55 <antonela> s/say/said
16:17:05 <irl> antonela: if i struggle i'll let you know
16:17:13 <antonela> irl, sounds fine, thanks!
16:18:00 <karsten> hmm, now it's gone?
16:18:16 <irl> it is part of 24422
16:18:24 <irl> it doesn't make sense to have two things
16:18:33 <karsten> ah!
16:18:34 <karsten> ok.
16:18:41 <karsten> great!
16:18:52 <karsten> then we have a few things we put under On Review.
16:19:09 <karsten> these are things that are, in theory, in progress but that are currently blocked by review.
16:19:24 <irl> the number seems to add up with the number i see in my trac query
16:19:33 <irl> i can have a reviews day tomorrow to catch up on these
16:19:50 <karsten> that would be cool!
16:20:00 <karsten> should we move these tickets to In Progress?
16:20:08 <irl> i think leave them there for now
16:20:13 <irl> so we remember to make sure they are unblocked
16:20:14 <karsten> we're not going to move them tomorrow anyway.
16:20:51 <karsten> I wonder how this works for the network team.
16:20:57 <gaba> do you change teh state of the ticket in trac to in review?
16:21:02 <karsten> yep.
16:21:04 <gaba> ok
16:21:24 <karsten> do you really have just 1 item per person in the In Progress column there?
16:22:04 <gaba> right now there are 2 for irl and 1 for you
16:22:30 <karsten> or should we move some of the ones in On Review to In Progress, assuming that I'm going to continue working on them as soon as they're unblocked?
16:22:44 <irl> ah ok, that's a good point
16:23:11 <gaba> mmm, i would move them once you are working on them
16:23:22 <karsten> okay.
16:23:38 <karsten> in that case I might start working on them tomorrow and have them done by next week.
16:23:44 <karsten> in which case we'd close them, not move them.
16:23:50 <gaba> ok
16:24:00 <karsten> works for me, too.
16:24:12 <karsten> we're fine tuning this thing as we go.
16:24:36 <karsten> okay, I think that's all for the roadmap.
16:24:42 <karsten> I guess.
16:24:54 <gaba> yep
16:25:05 <karsten> or are there urgent things in the backlog/icebox that should be In Progress?
16:25:26 <irl> even if there were i don't think we could do anything about it
16:25:34 <karsten> not much, that's true.
16:25:39 <karsten> okay, good.
16:25:53 <darkk> karsten: sorry, I was at IFF Glitter Meetup and missed a highlight. Thanks for reminder, I'll try to reply today (the aforementioned incident was finally resolved recently).
16:26:08 <karsten> ah, perfect!
16:26:13 <karsten> thank you. :)
16:26:22 <darkk> not yet :-)
16:26:27 <karsten> heh ok.
16:26:28 <karsten> ;)
16:26:46 <karsten> shall we move on to the next agenda item?
16:27:17 <irl> ok
16:27:18 <gaba> yes
16:27:29 <karsten> Rejected proposal and next steps (karsten)
16:27:40 <karsten> as you hear, our proposal was not accepted.
16:27:54 <karsten> we should briefly talk about next steps.
16:28:04 <gaba> yes
16:28:15 <gaba> Al is looking for other funding oportunities for it
16:28:16 <karsten> like: modify and try again, try elsewhere, try something different.
16:28:37 <karsten> as I wrote in that email, we'll have to do some of the work anyway.
16:28:53 <gaba> Yes. We can identify what needs to be done and still do it.
16:29:30 <karsten> so, should we wait for al to come up with other funding opportunities?
16:29:41 <karsten> and then decide how much we have to/want to change the proposal?
16:30:08 <gaba> yes, Al and the grants team is taking care of that part
16:30:10 <karsten> and in the meantime start the urgent part of the work anyway?
16:30:19 <irl> the problem here that we need to avoid is only doing enough to fix the current emergency and not enough to make it sustainable going forward
16:30:41 <karsten> well, we'd still do a rewrite.
16:31:02 <karsten> we'd just leave out the fancy parts at the end.
16:31:41 <karsten> should we discuss that plan in more detail next week?
16:31:50 <gaba> ok
16:31:50 <karsten> it seems that this week is already quite full.
16:31:51 <irl> ok yes, sounds good
16:32:52 <karsten> gaba: do I need to tell the grant team about this plan, or is it enough that you now know our plan? :)
16:32:56 <karsten> grants*
16:33:23 <gaba> I can talk with them about this plan. We do weekly meetings where I participate.
16:33:32 <karsten> okay, sounds good!
16:34:25 <karsten> next topic?
16:34:35 <irl> ok
16:34:42 <gaba> We are presenting projects for outreachy
16:34:46 <gaba> let me add this to the list
16:34:50 <gaba> yes, sorry, next topic
16:35:01 <karsten> outreachy?
16:35:23 <karsten> ok, on the agenda now.
16:35:35 <karsten> I might have to leave at 16:45 utc today. :/
16:35:37 <gaba> Pili is coordinating. You get an intern for 3 months for a specific project
16:35:43 <gaba> https://www.outreachy.org/
16:35:46 <gaba> focus on diversity
16:35:48 <pili> Hi!
16:35:57 <karsten> hi pili!
16:36:14 <karsten> but let's move on with the next topic, okay?
16:36:18 <gaba> I do not think we need to discuss anything today, right pili?
16:36:19 <gaba> yes
16:36:20 <pili> You have to commit 5h per week as a mentor I think
16:36:26 <gaba> please, sorry to interrupt in the middle of the agenda
16:36:30 <pili> I’ll let you move on
16:36:43 <karsten> can we discuss that next week? would that work for you?
16:36:58 <gaba> yes
16:37:09 <karsten> ok/sorry!
16:37:10 <gaba> I need to go soon too.
16:37:13 <karsten> tor-scaling@ discussion and next steps (karsten)
16:37:20 <karsten> gaba: ok.
16:37:27 <karsten> irl: did you follow that thread?
16:37:34 <karsten> the one thread on that mailing list.
16:37:41 <irl> yes, not in huge detail though
16:37:51 <karsten> ok. I'd be curious to hear your thoughts on that.
16:38:06 <karsten> also, we need to decide how much more to do.
16:38:16 <karsten> so far I looked at the failure/timeout/really slow cases.
16:38:30 <karsten> mostly because I think they affect user-perceived performance a lot.
16:38:35 <irl> i think i will have to read it in more detail, but then i can write up some thoughts
16:38:45 <karsten> what I didn't look into much is the slow cases.
16:38:56 <karsten> slow but not failure-slow cases.
16:39:20 <karsten> I could stop at this point and let the network team look into the failure cases first.
16:39:40 <karsten> I could do an initial analysis of the slow cases in the hope find obvious low-hanging fruit.
16:39:51 <karsten> or I could do a more thorough analysis of the slow cases.
16:39:56 <irl> i think for now we should see if the network team find what we have done useful
16:40:04 <karsten> as discussed earlier, we're at 6/3 points now.
16:40:24 <irl> if the network team thinks more analysis could be useful then we could target specifically what they want
16:40:36 <irl> but performing more analysis without that understanding could not be a good use of time
16:40:59 <karsten> I think we would be finding different issues than we found so far.
16:41:21 <karsten> example: I did find cases where tor sits there doing nothing for 15 or 35 minutes, and then bursts to completion.
16:41:52 <karsten> the slow cases are slow all the time, and the question is if we can do something about that. like avoiding relays that are slow or overloaded.
16:42:03 <karsten> I think the network team could do something with these results.
16:42:12 <karsten> but maybe they have enough to start with already.
16:42:33 <karsten> another aspect is whether we should start new measurements with existing onionperf right now.
16:42:47 <karsten> so that we have more data the next time we do an analysis.
16:42:52 <karsten> but I'm running out of time here.
16:43:01 <karsten> we should discuss this in more detail next week.
16:43:15 <karsten> for now, I'll focus on the things coming out of review tomorrow.
16:43:20 <irl> yes, let's discuss more next week
16:43:36 <karsten> ok.
16:43:49 <karsten> the rest can happen next week, too.
16:44:07 <karsten> should we end the meeting at this point and talk more next week?
16:44:17 <gaba> yes
16:44:20 <gaba> it sounds good
16:44:30 <gaba> I'm going to move the things we need to talk enxt week to an agenda next week
16:44:31 <karsten> okay!
16:44:36 <karsten> already did that.
16:44:38 <karsten> I think.
16:44:38 <gaba> oh, you did
16:44:39 <gaba> :)
16:44:40 <gaba> good
16:44:43 <karsten> alright, thanks, everyone!
16:44:48 <karsten> talk to you next week. bye! :)
16:44:51 <irl> bye!
16:44:53 <gaba> bye!
16:44:57 <karsten> #endmeeting