14:35:20 <karsten> #startmeeting metrics team meeting
14:35:20 <MeetBot> Meeting started Thu Apr 19 14:35:20 2018 UTC.  The chair is karsten. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:35:20 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:35:29 <karsten> * CollecTor/Onionoo backup operators: schedule (karsten)
14:35:42 <irl> if iwakeh is not yet here, maybe i could start with the snowflake topic?
14:35:49 <karsten> ah, or that.
14:35:59 <irl> * Number of snowflake embeds from webstats (irl)
14:36:29 <irl> we haven't been asked for this yet, but it seems like something that we might be asked for and i wanted to get an idea of how much work it would be to implement
14:36:31 <karsten> hi iwakeh!
14:36:36 <iwakeh> hi there!
14:36:42 <karsten> irl: okay. can you explain?
14:36:51 <karsten> iwakeh: we just started.
14:37:14 <irl> basically, it's a matter of looking through the webstats for snowflake.torproject.org (which is a static site) and seeing how many hits there are on /embed.html that would indicate that the embed.html code has been loaded when embedded somewhere
14:37:30 <irl> we might want to subtract the demo page from those stats which would be /index.html (or we may not)
14:37:39 <karsten> sounds doable.
14:37:44 <karsten> we'd need those logs.
14:37:47 <irl> we have the logs
14:37:50 <irl> they're already in collector
14:37:54 <karsten> aha!
14:37:56 <karsten> since when?
14:38:05 <irl> i guess they just appear when new static sites are added
14:38:06 <karsten> (haven't seen that virtual host.)
14:38:23 * iwakeh checking ...
14:38:35 <irl> it's a new virtual host, moved from the previous server that wasn't a tpo server
14:38:42 <karsten> so, the idea would be to add a graph?
14:38:52 <irl> yeah, how many times has that page been loaded
14:39:04 <iwakeh> for example:https://collector.torproject.org/recent/webstats/snowflake.torproject.org_aroides.torproject.org_access.log_20180407.xz
14:39:09 <karsten> and it's okay that we don't have query parameters?
14:39:16 <karsten> just the hit to that page?
14:39:29 <irl> yep, it would be a number you could compare to stats that we later get from the distributor
14:39:43 <karsten> not much work.
14:39:54 <irl> ok awesome (: i'll report this back and see if we should do it then
14:39:55 <karsten> hardest part for me would be to write a useful description.
14:40:06 <karsten> because I don't know much about snowflake.
14:40:26 <karsten> so, I'd appreciate input on the text part. let me give you a link..
14:40:32 <irl> i think this would be for the snowflake person to come up with, or i could give it a go and someone could double check
14:40:55 <karsten> https://gitweb.torproject.org/metrics-web.git/tree/src/main/resources/web/json/metrics.json
14:41:03 <karsten> we'd need the description part.
14:41:24 <irl> ok cool, and that would be the hardest part?
14:41:31 <karsten> I guess so.
14:41:45 <irl> ok, i'll report back then and see if this is a useful thing to do.
14:41:56 <karsten> sounds good! say it's going to be super hard. ;)
14:42:00 <karsten> but we'll do it!
14:42:01 <irl> hehe (:
14:42:18 <karsten> okay, great.
14:42:20 <karsten> next?
14:42:34 <irl> ok
14:42:38 <karsten> * CollecTor/Onionoo backup operators: schedule (karsten)
14:42:55 <karsten> this is related to iwakeh's suggestion.
14:43:12 <iwakeh> Hand over very soon.
14:43:14 <karsten> suggestion to start early with handing over operation.
14:43:19 <karsten> yes.
14:43:29 <iwakeh> If your time permits, irl?
14:43:41 <iwakeh> with me as backup.
14:44:19 <irl> i'm doing a lot of handover at work currently, so i don't think we can start anything this week. probably next thursday afternoon would be a good time to start with this.
14:44:51 <karsten> thursday afternoon is a bit tricky here, with meetings.
14:44:53 <iwakeh> Do you have access to everything already?
14:44:58 <karsten> yes, he does.
14:44:59 <irl> i do already have access to everything
14:45:06 <karsten> we already looked at all the hosts.
14:45:07 <iwakeh> good.
14:45:26 <karsten> like, right after this meeting I have another one.
14:45:27 <iwakeh> well, what else needs to be transferred?
14:45:35 <iwakeh> regarding knowledge?
14:45:50 <irl> i guess, procedures for common tasks
14:45:59 <karsten> maybe we could do the next deployment together.
14:46:07 <irl> yeah, that would be good
14:46:07 <iwakeh> for example
14:46:23 <iwakeh> There are hardly any procedures.
14:47:02 <karsten> so, this depends on when we put out the next release then.
14:47:04 <iwakeh> when's the next release?
14:47:07 <karsten> heh
14:47:13 <iwakeh> oh
14:47:20 <iwakeh> Another thing:
14:47:28 <iwakeh> release-dance!
14:47:33 <karsten> yep.
14:48:09 <karsten> so, sounds like we can't schedule this other than putting out new releases.
14:48:37 <karsten> and with the next topic in mind, it might be that we're not putting out a new release in the next week.
14:48:39 <iwakeh> metrics-lib 2.3.0 updates
14:49:22 <iwakeh> ?
14:49:32 <karsten> hmmmmm
14:49:46 <karsten> we still have many weeks to do this.
14:49:54 <iwakeh> true, no rush.
14:50:04 <iwakeh> My suggestion was
14:50:07 <karsten> maybe we should rather wait for some real changes to pile up before releasing them.
14:50:25 <karsten> where wait == work on
14:50:27 <iwakeh> really tight to the first requirement, and that is
14:50:38 <iwakeh> irl's time for doing these things.
14:50:44 <iwakeh> Other than that
14:50:47 <karsten> yep.
14:50:52 <iwakeh> just as soon as possible.
14:50:57 <irl> if mostly it involves deploying new releases, then maybe there's not so much to do
14:51:14 <karsten> there's not.
14:51:21 <iwakeh> true
14:51:38 <karsten> sounds like a plan.
14:51:38 <irl> ok, so i'm less worried about this now, we should just make sure to have releases in the next month or so
14:51:49 <karsten> yes, that sounds doable.
14:51:53 <iwakeh> fine
14:52:11 <karsten> great!
14:52:53 <karsten> next?
14:53:03 <irl> ok
14:53:21 <karsten> * Sponsor 13 priority (karsten)
14:53:32 <iwakeh> extended?
14:53:40 <karsten> no response yet.
14:53:52 <karsten> this is mostly about making progress this month.
14:53:57 <karsten> to show that we're working on this.
14:54:07 <iwakeh> sure, I plan to have
14:54:14 <karsten> I'm going to focus more on this in the remaining 1.5 weeks of this month.
14:54:22 <karsten> this == deliverable 2.
14:54:29 <iwakeh> a structure for deliverable 1 next week.
14:54:34 <karsten> awesome!
14:54:58 <karsten> I'll incorporate feedback we got.
14:55:06 <karsten> the question is, what should I do next?
14:55:14 <karsten> ask other folks for more feedback?
14:55:23 <karsten> write other parts (bridge users)?
14:55:39 <karsten> look into formatting formulas?
14:55:53 <iwakeh> that would be another section == progress!
14:56:12 <karsten> that == bridge users?
14:56:17 <iwakeh> and also progress on form(at)!
14:56:23 <iwakeh> yes
14:56:34 <karsten> okay, so those two things.
14:56:44 <karsten> btw, which api did you suggest again?
14:56:51 <karsten> for math formulas.
14:57:02 <iwakeh> MathJax
14:57:05 * irl suggested mathjax too
14:57:07 <karsten> irl and I very briefly talked about this topic.
14:57:20 <karsten> without looking, what's the plan for js-free clients?
14:57:24 <iwakeh> that's sort of standard
14:57:48 <iwakeh> oh, good old ascii formulas :-)
14:58:25 <iwakeh> see sage-math for example, it does LaTeX and good old text.
14:58:42 <karsten> ah ok. so, we write it once, it produces two things?
14:58:44 <irl> i would prefer that we have LaTeX and then just render them as PNGs for a non-JS version
14:58:46 <iwakeh> as output.
14:58:57 <irl> LaTeX -> MathJax and PNG is easy
14:59:08 <iwakeh> sage-math is not for printing.
14:59:20 <iwakeh> I think irl's suggestion is best.
14:59:49 <iwakeh> then we could even skip mathjax.
14:59:56 <karsten> ah.
14:59:57 <irl> Wikipedia takes LaTeX and gives you a PNG
15:00:00 <iwakeh> LaTeX -> html works too.
15:00:01 <irl> when editing articles
15:00:31 <karsten> hmm, it sounds I'm not the best person to do that part.
15:00:39 <irl> you can always add alt attributes for screen readers
15:00:45 <karsten> maybe I'll focus on content for now.
15:00:48 <karsten> to make more progress.
15:01:10 <karsten> I mean, I can write LaTeX formulas.
15:01:16 <iwakeh> yes, form follows function (or content in our case here).
15:02:03 <iwakeh> LateX format will be malleable to whatever we choose in the end.
15:02:30 <iwakeh> So, maybe, write using LaTeX and decide later about how to put that on the web?
15:02:39 <karsten> ah, well,
15:02:48 <karsten> we already have things in html.
15:02:53 <karsten> I meant the formulas.
15:03:13 <karsten> suggestion: I write them in ascii for now, and we decide this part later.
15:03:23 <karsten> happy to convert ascii to latex.
15:03:26 <karsten> or something else.
15:03:26 <iwakeh> Or simply put placeholders for formulas for now?
15:03:30 <karsten> yes, or that.
15:03:48 <iwakeh> that keeps all options open.
15:03:59 <karsten> ok.
15:04:19 <karsten> alright!
15:04:35 <karsten> I think that's all for today. unless there's anything else?
15:04:44 <iwakeh> all done, I think.
15:05:16 <irl> all done
15:05:26 <karsten> great! thanks! bye, bye. :)
15:05:30 <iwakeh> bye, bye!
15:05:33 <irl> bye!
15:05:35 <karsten> #endmeeting