Friday, 2018-02-09

*** diablo_rojo has joined #openstack-tc00:14
*** harlowja has joined #openstack-tc00:57
*** kumarmn has joined #openstack-tc01:06
*** kumarmn has quit IRC01:11
*** kumarmn has joined #openstack-tc01:21
*** kumarmn has quit IRC01:52
*** liujiong has joined #openstack-tc02:30
*** mriedem has quit IRC02:44
*** ykarel|away has joined #openstack-tc03:44
*** ykarel|away is now known as ykarel03:45
*** liujiong has quit IRC03:46
*** diablo_rojo has quit IRC03:47
*** rosmaita has quit IRC04:02
*** harlowja has quit IRC05:10
*** coolsvap has joined #openstack-tc05:11
*** dtantsur|afk is now known as dtantsur08:08
*** jpich has joined #openstack-tc09:02
*** ykarel is now known as ykarel|lunch09:08
*** ykarel|lunch is now known as ykarel10:11
*** openstackgerrit has joined #openstack-tc10:16
openstackgerritMerged openstack/governance master: Replace Chinese quotes to English quotes  https://review.openstack.org/54119710:16
openstackgerritMerged openstack/governance master: Rename goals/template.txt to .rst  https://review.openstack.org/54149610:17
openstackgerritMerged openstack/governance master: Remove editor mode setting from goals template  https://review.openstack.org/54149010:17
*** dtantsur is now known as dtantsur|bbl10:43
*** chandankumar has quit IRC10:54
*** cdent has joined #openstack-tc11:17
*** chandankumar has joined #openstack-tc11:17
*** cdent has quit IRC12:31
*** dtantsur|bbl is now known as dtantsur12:40
*** cdent has joined #openstack-tc12:57
*** fdegir has quit IRC13:14
*** fdegir has joined #openstack-tc13:15
*** bauzas is now known as bauwser13:45
*** cdent has quit IRC13:53
*** kumarmn has joined #openstack-tc13:54
*** mriedem has joined #openstack-tc13:57
*** kumarmn has quit IRC13:58
*** cdent has joined #openstack-tc14:01
*** TheJulia has quit IRC14:02
*** TheJulia has joined #openstack-tc14:03
*** johnsom has quit IRC14:03
*** rosmaita has joined #openstack-tc14:03
*** johnsom has joined #openstack-tc14:04
*** kumarmn has joined #openstack-tc14:28
*** david-lyle has quit IRC14:29
*** coolsvap has quit IRC14:33
*** kmalloc has quit IRC14:40
*** kmalloc has joined #openstack-tc14:41
*** hogepodge has quit IRC14:51
*** hogepodge has joined #openstack-tc14:52
*** ykarel is now known as ykarel|away14:56
*** ykarel|away has quit IRC15:01
*** hongbin has joined #openstack-tc15:06
openstackgerritMerged openstack/governance master: Add devstack-plugin-container to QA team  https://review.openstack.org/54090615:14
*** david-lyle has joined #openstack-tc15:30
*** david-lyle has quit IRC15:51
*** david-lyle has joined #openstack-tc15:55
*** cdent has quit IRC16:06
*** jpich has quit IRC17:32
*** harlowja has joined #openstack-tc17:37
*** david-lyle has quit IRC17:50
*** dtantsur is now known as dtantsur|afk18:45
*** david-lyle has joined #openstack-tc19:02
*** mriedem is now known as mriedem_afk19:06
*** lbragstad has quit IRC19:18
* persia decides not to publish an article entitled "how to collude with your business partners and subvert the openstack bureaucracy", but a recent conversation has suggested that some of the concepts that might be in such an article may be useful for discussion towards enhancing some of the developer and developer-sponsor onboarding guides, either at some office hours, or at the PTG.19:22
*** cdent has joined #openstack-tc19:28
* fungi is entirely in favor of subverting bureaucracy if it leads to better free software19:29
fungipeace, love and free software, man19:30
persiaTechnically, we have very little bureaucracy, as very few of our roles are appointed (vs elected), but the perception is certainly present.19:30
*** cdent has quit IRC19:55
dimsw00t fungi20:12
dmsimardpersia: the topic of OpenStack summit presentation voting came up recently, I reminded them that we were once considering removing them entirely and why it was a problem20:19
persiaI'm still not delighted with the implementation, but it is better than it has been.  But I don't like any process that I can use to unfair advantage (which I haven't done for OpenStack Summit speaker selection since the last time we were in Vancouver, but still)20:20
dmsimardRemoving the direct links to presentations for voting had the awkward side effect of people continuing to advertise in different media/channels but with like "search for X and vote for me!"20:21
persiaYep.  When using voting, one needs to provide media for political statement, or one ends up with segmented audiences.20:22
fungisure, technical solutions to social problems rarely work completely20:22
persiaAlso, democracy rarely works completely.  My preferred model for talk selection is elected "track chairs", each of whom curates a track based on their expertise.  Using "track committees" with 3 or more members instead of a single official makes it harder to subvert (and leads to better tracks).20:24
fungii still think we need to drop the session voting altogether. the times i was a track chair the voting results rarely provided useful insights20:24
dmsimardAre track chairs elected right now ? or appointed ? I guess I'm not familiar with the way it works behind the scenes.20:25
cmurphyagreed, the votes rarely make a difference in selection20:25
fungii'd be in favor of voting for track chairs instead20:25
fungidmsimard: appointed volunteers20:25
persiadmsimard: Appointed, via a process that is subject to adjustment by interested parties.20:26
fungithe foundation attempts to balance chairs across different roles and demographics for some semblance of representative/diverse opinions20:26
cmurphyi don't think i'd be in favor of voting for track chairs, it would be easy to stuff the ballots there20:26
fungibut to a great extent the track chair selection is determined by who volunteers to do the work20:26
fungicmurphy: yeah, it does become harder to ensure a balanced representation20:27
persiaA technical solution that I have seen work for smaller groups is to give attendees a massive menu with "I would attend" indications.  Then calculate an optimal schedule so that the most people can attend the most talks they want with the least conflicts.  The problems with that model are 1) people lie about attendance, 2) the computation is difficult, 3) events are paid for by sponsors who like to have influence20:27
dmsimardpersia: I think that's the approach we've had in past summits.. like a scale from 1 to 5, 5 being "I would attend this for sure"20:28
*** kumarmn has quit IRC20:28
fungibut to a great extent the event organizers have a feel for who will be in attendance and the goal is to find people who will pick the right mix of content for the attendee demographics20:28
persiadmsimard: A scale prohibits a scheduling algorithm.20:28
dmsimardpersia: right -- it has to be binary, and even that isn't perfect because of your #1. As a meetup organizer, ordering the right amount of pizzas based on the amount of signed up folks is quite challenging :)20:29
dmsimardSometimes it can be as bad as 50% no show20:29
persiadmsimard: Some conferences embed RFID tags in badges and sensors at doors, and then use the results to decide whether to trust people next time.  Only works for communities with significant repeat attendees though.20:30
dmsimardI suppose for paid conferences the rate is not that high, though20:30
dmsimardpersia: interesting logistics there :)20:30
fungithe challenge can be similar though. this is part of the reason the registration rates increase over the registration period20:30
persiaPaid conferences mean people get badges.  This has very little correlation with talk attendance.  Lots of folk pay for conferences to be able to be in the halls, or to have an excuse to be in the city and meet someone else in a hotel room for 20 minutes.20:31
fungiincentive to register early, since more registrants early allows for more accurate content and planning20:31
persiaWhat incentive matters though?  There are lots of price-insensitive folk who attend paid conferences, especially business-focused conferences.20:32
persiaGiven the cost of travel + 1 week FTE + lodging, the difference between $0 and $2000 is often not even considered.20:32
dmsimardNo real incentive other than prices, flight, early-bird and hotel availability etc20:32
persiaPrices really don't matter except for individual developers.20:33
fungiyep. the early bird pricing model is just one more tool to try and get some early demographic info, on the assumption that some information is better than none20:33
fungii agree it's still skewed and i'm sure they take that into account20:34
*** kumarmn has joined #openstack-tc21:02
*** kumarmn has quit IRC21:21
*** kumarmn has joined #openstack-tc21:21
*** cdent has joined #openstack-tc21:25
*** cdent has quit IRC21:28
*** DuncanT has quit IRC21:49
*** DuncanT has joined #openstack-tc21:49
*** kong has quit IRC22:05
*** kong has joined #openstack-tc22:06
*** ildikov has quit IRC22:13
*** ildikov has joined #openstack-tc22:14
*** mriedem_afk has quit IRC23:16

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!