Thursday, 2019-05-09

*** jamesmcarthur has joined #openstack-tc00:04
*** jamesmcarthur has quit IRC00:05
*** jamesmcarthur has joined #openstack-tc00:05
*** mriedem has quit IRC00:15
openstackgerritZane Bitter proposed openstack/governance master: Add Python3 update goal for Train  https://review.opendev.org/65790800:30
*** Sundar has quit IRC00:45
*** jamesmcarthur has quit IRC01:12
*** dangtrinhnt has joined #openstack-tc01:18
*** jamesmcarthur has joined #openstack-tc01:43
*** jamesmcarthur has quit IRC01:44
*** jamesmcarthur_ has joined #openstack-tc01:44
*** jamesmcarthur_ has quit IRC02:33
*** jamesmcarthur has joined #openstack-tc02:34
*** ricolin has joined #openstack-tc02:42
*** jamesmcarthur has quit IRC03:08
*** jamesmcarthur has joined #openstack-tc03:11
*** jamesmcarthur has quit IRC03:34
*** e0ne has joined #openstack-tc03:47
*** jamesmcarthur has joined #openstack-tc04:05
*** e0ne has quit IRC04:46
*** samueldmq has quit IRC05:19
*** sarob has joined #openstack-tc06:06
*** sarob has quit IRC06:11
*** whoami-rajat has joined #openstack-tc06:20
*** Luzi has joined #openstack-tc06:26
*** sarob has joined #openstack-tc06:46
*** sarob has quit IRC07:20
asettleMooorning o/07:50
*** jpich has joined #openstack-tc07:58
*** e0ne has joined #openstack-tc08:14
*** e0ne has quit IRC08:17
*** gmann_pto has joined #openstack-tc08:59
ttxRe: removing https://wiki.openstack.org/wiki/OpenStack_health_tracker -- it was mentioned as one of the rare places ops could use to assess maturity in https://etherpad.openstack.org/p/DEN-Operator-end-user-Public-cloud-feedback09:16
ttx(while health != maturity, I think we might want to address their need somewhere)09:17
*** wxy-xiyuan has quit IRC09:26
*** e0ne has joined #openstack-tc09:38
*** e0ne has quit IRC09:40
mugsiematurity is not easily measured - for some applications a project may be very mature, while for others, lacking. capturing that is very difficult, and really needs prospective deployers to do some background research09:44
ttxmugsie: totally agree, just mentioning that we'll be soon removing something they apparently used as part of that research09:55
*** e0ne has joined #openstack-tc09:56
mugsieyeah, I wonder how we can help them move away from relying on things like ^09:59
*** e0ne has quit IRC09:59
ttxMaturity is really in the eye of the beholder. There is a reason why the adoption curve exists.10:02
ttxSo beyond showing off users and % of adoption, not sure of what we can do really (beyond identifying the things that do not work at all and removing them)10:03
mugsieyeh10:03
*** mugsie has quit IRC10:25
*** mugsie has joined #openstack-tc10:35
*** mugsie has quit IRC10:35
*** mugsie has joined #openstack-tc10:36
*** mugsie has quit IRC10:38
*** mugsie has joined #openstack-tc10:39
smcginnisSounds like the project's fears that that health tracker would be interpreted in a way different than what it was intended were well founded.11:08
asettleWe're smort like that, smcginnis ;)11:24
mugsieor lucky :P11:30
mugsieasettle: at some point today / tomorrow, I am going to try and drop a review on each service with the vision retrospective stuff, and stick it in an etherpad. do you want to send that email after I do that and link to the etherpad?11:31
*** gmann_pto has quit IRC11:50
asettleYep that sounds dandy. I only just re-started work today, so I'm in massive catch-up mode. The email would be sent tomorrow if anything. I'm fighting SAP right now.12:00
mugsieheh - I know the feeling.12:05
asettleEvery now and then it just *freezes*12:05
asettlelike.12:05
asettleI only have so much patience? And it's somewhat limited.12:05
mugsieI am surpised there is any left after a week in a PTG :P12:06
asettleI took two days off when I got home12:07
asettleI thought I'd be productive but all I ended up doing was sawing the cord for the hedge trimmer in half about 10 minutes into my hedge trimming...12:08
*** jamesmcarthur has quit IRC12:21
*** jamesmcarthur has joined #openstack-tc12:21
*** jamesmcarthur has quit IRC12:30
ttxtc-members: One thing that we asked during PTG was ability to add our own TC questions to the user survey. I reached out to OSF staff in charge and that is definitely possible (think 1-2 questions per survey), but the new user survey is being publicized really soon now, so if we want to add questions for this one it would be great to come up with our questions before.... early next week or so12:41
ttxCan't really remember what we had in mind specifically but maybe someone remembers12:41
fungione thing which came up was finding out what openstack services folks are deploying12:43
*** jamesmcarthur has joined #openstack-tc12:49
asettleThat was the main one, we also had an item about getting SIGs into the UC. But I'm not entirely too sure what the  context was12:50
asettleAll the note says is "Reach out to OSF/UC to increase contentquestion in user survey!12:50
asettle"And it's under the Action items for SIG governance12:50
asettleI wonder if ricolin remembers12:50
*** ijolliffe has joined #openstack-tc12:50
*** mriedem has joined #openstack-tc12:51
ttxfungi: I think we already have that question (which components are you deploying)12:55
fungiyep, i thought so too12:56
asettleSo I think the question wasn't answered if we did. Are we getting the right info from the way it is phrased?12:56
fungibut the fact that it came up as something we wanted the survey to address made me question my memory there12:56
asettleIt was to do wit hthe evolution of the help-most-needed list. Hoping to get lists of companies associated with projects who may need more help.12:57
asettleSo we would have the opportunity to reach out to them personally. Using Allison's idea12:57
ttxasettle: ah yes, so that was more about getting access to more data than the one showed in reports. That would likely be on a case-by-case basis due to anonymity requirements, but yes12:59
fungioh, right, we wanted to know what projects a given respondent was contributing to (if any)13:00
fungiso we could try to match those up to the projects they were using13:00
asettleYes, all that. I think we need to find out if we can get access, ttx ? And if so, what can we really do with it.13:01
fungiand see whether there was a correlation13:01
asettleBut also an option if we could potentially add a question to the UC about contributing? Or perhaps something like "Are you aware of the help most needed list?"13:01
*** jamesmcarthur has quit IRC13:05
*** samueldmq has joined #openstack-tc13:16
persiaAsking folk "are you aware of this" in the user survey is probably not ideal, as there are a limited number of questions and respondants have a limited amount of patience.13:22
persia"To which projects do you contribute" is an interesting question, and lots of good data mining can happen from that.13:23
persiaProbably worth looking over the questions on the *last* survey as input on which questions on the *next* survey would help explain answers of interest.13:24
fungisince the user survey tends to be about a few people responding on behalf of others in their organization we might instead want to ask to which projects their organization is contributing, rather than the just projects contributed to by the individual respondent13:26
mugsie^^13:27
dhellmannwe can find the organizations contributing to a project by looking at gerrit, but it might still be interesting to ask it as a way to nudge them to thinking about contributing to what they use13:27
dhellmannesp. if the contributing question comes immediately after the using question13:28
*** lbragstad_ has joined #openstack-tc13:29
*** lbragstad_ is now known as lbragstad13:30
asettleGood suggestion fungi13:34
asettleI'm just thinking out loud :)13:34
asettledhellmann, yes exactly13:34
mugsiedot13:34
mugsiegha13:35
asettle... u k there13:36
smcginnismugsie: Making some drawings? :)13:37
mugsieyea - damn terraform is complex13:37
ttxSo the main issue with accessing the "which company uses what" data is that it is submitted under a privacy promise -- TC members could access that data if they are OK to sign a confidentiality agreement (the UC members sign it), but we would not be able to openly discuss that data afterwards13:38
asettleDoes that kinda... defeat the purpose of us gathering that data? We would, in theory, need to discuss the data found to be able to approach companies?13:39
ttxso it might be difficult to make it a "TC thing"13:39
ttxOur current approach is that we/OSF act as a proxy to contact the relevant companies in corner cases13:40
asettleSo, in theory OSF could?13:40
ttxasettle: without the privacy promise we would get much less data and much more imcomplete statements13:40
ttxasettle: the OSF does. I did contact Ceilometer users when we had the leaderless situation lately13:40
asettleThat's fair. I'm not suggesting we remove it, I'm just saying that us signing the privacy agreement would mean we'd not actually able to fulfil our promise/idea/plan13:41
asettleIn theory, would we able to read the data and discuss with OSF?13:41
ttxyes13:41
asettleAhhh13:41
asettleOkay13:41
asettleThat changes it13:41
ttxTC members can access that data if they are OK with signing a confidentiality agreement, which would ask that they not discuss that data openly13:42
ttxIt's a thin line to walk on -- we get a lot of answers based on the fact that this data is confidential13:43
*** altlogbot_3 has quit IRC13:43
ttxUser survey has a "is it OK to talk about your usage publicly" checkbox and not so many check that13:44
*** altlogbot_1 has joined #openstack-tc13:45
dhellmannperhaps we want just 1-2 people to agree to sign so they can assist wendar contact those companies13:46
*** redrobot has quit IRC13:46
ttxyeah, that would be good. FWIW fungi and me can probably already access that data, but if anyone else wants to sign things...13:48
*** Guest89536 has joined #openstack-tc13:49
persiaOn data mining: it would be interesting to have summary data public, even if the specifics can't be.  Things like: for each project, what percentage of consuming organisations are contributing organisations?  Similarly, regression analysis to determine if there are interesting correlations between usage of and/or contribution to various projects and contribution to other projects.13:49
persiaNone of that should violate privacy, and I expect that for TC purposes, the names of the individual organisations don't matter that much.13:50
asettlettx, maybe we just keep it at that. Would you and fungi be comfortable taking on that role that rides that line for us? Saves it getting potentially messy.13:50
persiaasettle: It's probably worth considering the possibility that there may not be an overlap between foundation staff and TC members in the future.  Doesn't require anything now, but does require consideration later if elections go a different way.13:51
dhellmannif we're going to try to do serious analysis of contributions, we need to be very clear about what a contribution is. because otherwise asking someone to self-report if they contribute is going to give us answers all over the spectrum.13:51
asettlepersia, sure - but let's not over complicate the now. ttx will at least be in for another year. So that's a start, and once/if anything changes, we can reassess.13:52
asettledhellmann, good point. Definitions are important here.13:52
persiaI'm not sure that's a bad thing.  We have the actual data of who contributes to what and how (or nearly so: affiliation is a tricky beast).  Understanding how that is viewed is likely to help determine appropriate future messaging.13:52
*** Guest89536 is now known as redrobot13:55
fungisorry, catching back up on the conversation. i don't think the tc wants a list of which companies are using and contributing to what, right? instead we want to know whether projects are seeing contributions from the people who are using them (and also provide some nudging via this question to remind those taking the survey that's how open source works?)13:57
fungiso the responses don't need to be public, the analysis can be anonymized sufficiently13:58
dhellmannfungi : yes, that's what I thought we wanted. I don't think someone answering yes because the file bug reports is the sort of contribution we want them to be considering, though.13:58
asettleYes - that's right fungi13:58
dhellmann*they file13:58
fungisure, we can word it in such a way as to not indicate that filing bugs is what we're asking about13:58
ttxyeah, I guess that data could be produced and discussed using anonymized data14:00
ttx"Company Z"14:00
persiaIn conversation with various folk, I've heard the case that folk don't file bug reports, and instead just decide they don't like OpenStack.  Let's not let our desire for organisations to contribute full-time development staff blind us to the entire range of useful contribution that lets us succeed.14:01
ttxalthough it would probably be trivial to deduce who is what14:01
ttxbrb14:01
persiattx: Or just ID numbers for companies, with no names.14:01
persia(and if we only have that data for a few of the questions, it is probably harder to deduce who is what)14:01
dhellmannif we only looked at the survey answers, yes, but we can get a list of contributing companies by looking at patches and reviews in gerrit14:02
persiaOh, heh, yeah, I suppose if we have different lists of "these are the projects to which our org contirbutes", and mine gerrit, except for apparent duplicates, we get all the names.14:03
fungiby "anonymized" i really meant "sufficiently generalized such that we can draw the conclusions we need with no ability to identify individual responses"14:15
zanebI'd be interested in a checklist about how you contribute to OpenStack that includes bug reports, contributing code, reviewing patches/maintenance, docs14:16
fungiso in other words define the analysis we wish performed and have someone with access to the data perform that analysis and provide the results14:16
persiazaneb: Maybe two questions: one asking for the list of projects and the second asking the types of contributions?14:24
zanebyes, they'd need to be separate14:24
zanebI like dhellmann's push-polling idea too ;)14:24
dhellmannwould we want them separate, or as 1 big matrix with contribution types for each project? I guess that might be asking for too much14:25
zanebyeah, I think that's asking too much14:26
dhellmannwe could probably deduce bug reporting affiliation for each project, too, the tools I built just don't do that14:26
zanebtbh we know what projects companies are contributing to because contributions are public. the foundation could correlate that data already14:27
zanebif we're going to add a question it's more to get users to think about that for themselves14:27
dhellmannyeah14:32
ricolinasettle, the `Reach out to OSF/UC to increase contentquestion in user survey!` is for asking again on how can we put SIGs' user question into user survey or anyother format14:42
asettleAh right, sure. Thanks.14:50
*** Luzi has quit IRC14:51
lbragstado/14:59
zaneboh hey we have office hours15:00
asettleWaddduppp15:00
ttxyo15:02
ttxreiterating due to urgency:15:03
ttxtc-members: One thing that we asked during PTG was ability to add our own TC questions to the user survey. I reached out to OSF staff in charge and that is definitely possible (think 1-2 questions per survey), but the new user survey is being publicized really soon now, so if we want to add questions for this one it would be great to come up with our questions before.... early next week or so15:04
ttxIf all we want is the anonymized contrib/company metrics, we don;t really need extra questions15:04
dhellmannttx: do you have access to the wording for the current question about which projects are used?15:05
dhellmannbrainstorming question wording in https://etherpad.openstack.org/p/tc-user-survey-201915:05
ttxLet me ask aprice15:08
ttxIIRC people describe (potentially several) deployments and then check boxes15:08
dhellmannk15:09
ttx"Which projects does this deployment currently use, or are you interested in using in the future?"15:09
dhellmannok15:09
ttxthis provides three options: production, testing and interested15:09
ttxthat is the question if the respondent indicates their deployment is in production15:09
ttxif it's a POC, the options are testing and interested15:10
ttxok thanks!15:10
ttxoops wrong window15:11
asettleClassic.15:11
ttx"how would you describe your job ?15:11
ttx"15:11
ttx"I copypaste others responses between windows"15:11
dhellmannI write email.15:11
ttxdhellmann: that is next level shit!15:12
ttxI read email15:12
dhellmann10x contributor15:12
dhellmannsomeone with a less muddled brain and better language skills please review https://etherpad.openstack.org/p/tc-user-survey-201915:12
ttxif you have more questions, aprice is listening and can save me teh copypaste now15:12
aprice:)15:12
ttxI /think/ we tend to limit changes to wording because they introduce a skew in trending data15:13
dhellmannaprice, your input on https://etherpad.openstack.org/p/tc-user-survey-2019 would be great, those are just to get the ball rolling15:13
ttx(i.e. ability to compare results year over year)15:13
apricewording changes to existing questions are also hard because of the effects on translation15:13
dhellmannyeah, I was hoping to phrase this new contributing question closely to the using question, rather than change anything old15:13
*** jamesmcarthur has joined #openstack-tc15:13
ttxdhellmann: ah, additional question? That would definitely be fine15:14
ttxdhellmann: I like your wording fwiw15:14
apricedhellmann: i like the wording. on the second question, I suggested adding "Other" with a text box to see what users contribute that we may not be proactively listing15:16
dhellmann++15:16
apriceon the second question, im getting a little stuck because the sponsoring of events is something that someone's organization would do, but the others are up to the individual. i proposed an edit15:18
dhellmannyeah, that's a good point15:18
dhellmannmaybe the question should be "how do  members of your organization contribute..."?15:18
dhellmannrather than "how do you..."?15:18
apriceyeah i like that15:19
mugsieyeah, "members of you org" is a good qualifier15:19
dhellmanndo we want to list the PTG separately?15:20
dhellmannI reordered a bit to pull all of the in-person stuff together15:21
dhellmanndo we want to list things like third-party CI, release management, and infra?15:22
asettleConsidering we include those as teams in our docs, yes?15:23
dhellmannah, that was specifically about question 215:23
dhellmannmaybe including release and infra in question 1 is sufficient15:24
* lbragstad makes a first pass at https://www.lbragstad.com/blog/openinfrastructure-summit-denver-technical-committee-summary15:26
* dhellmann adds lbragstad's post to his reading queue15:27
evrardjp"All of the above"15:33
apricedhellmann: lmk when the questions are final, and I can work with our team on getting them uploaded to the survey. we are hoping to have our big promotion push starting soon and this cycle will close at the end of August.15:38
dhellmannI wonder if we want to ask something like "if you do not contribute maintenance resources, why not?"15:38
dhellmannaprice : sounds good. I expect mnaser will want to look it over so he's at least aware of the questions15:39
apricedhellmann: sounds good - feel free to ping me here if y'all need anything else or have any questions15:39
dhellmannthanks, aprice!15:40
asettleThis looks good dhellmann and aprice15:41
dhellmannI'm not happy with line 10 yet15:41
dhellmann"What obstacles to contribution have you encountered?"15:42
dhellmannthat presumes they've tried, though15:42
asettleYou could go simple like...15:42
asettle"If you do not contribute any maintenance resources to a project, please explain why:"15:42
asettle[text box}15:42
dhellmannyeah, I was trying to be a little more indirect with the why bit15:43
mugsie"you are bad and should feel bad"15:43
asettleMaybe not "please explain why" cause it's kinda aggressive.15:43
asettleHAHAHAHA15:43
asettlemugsie, in summary15:43
dhellmann"we assume you want to, what is making it hard for you?"15:43
asettleYeahhh but sometimes direct is best. If I was to read "what is the main obstacle" as a user, I might just be like :shrug:15:43
asettleI get why we don't want to be aggressive in a survey though15:44
asettleJust a note15:44
dhellmannhow's that?15:45
asettle\ye15:45
openstackgerritColleen Murphy proposed openstack/governance master: Add caching as a base service  https://review.opendev.org/65814215:45
mugsiedhellmann: ++15:46
zanebmugsie: ROFL15:47
dhellmannI'm happy with these. If we don't get good answers, then next year we can as "which project would we have to deprecate to get you to start contributing?"15:49
zanebha15:50
asettleHahhahah great15:50
apriceso do we need to include a "None" option for question 1 that would prompt line 12?15:50
*** timburke has joined #openstack-tc15:51
zanebdhellmann: maybe ask what prevents you from contributing *more*?15:51
dhellmannaprice : I think with a "check all that apply" note having none selected would be ok? unless "None" is easier to tally somehow?15:51
dhellmannwe do also want to know what might be preventing someone who does contribute from contributing *more* so I tried to phrase it that way15:52
ttxzaneb: I like that, less aggressive, same results15:52
dhellmannzaneb : I tried for that, but suggestions on wording welcome15:52
zanebI'm suggesting we literally just shove the word 'more' between 'contributing' and 'maintenance resources'15:53
dhellmannmaybe "or makes it difficult to contribute more?" at the end of line 12?15:53
apriceah ok. i wonder if it could event be something around barriers instead of preventing? because preventing makes it sound like they dont contribute at all15:53
dhellmannah, yeah, that simple edit works too15:53
zanebadded it in. feel free to rework15:53
apricei like that15:54
dhellmannI like the "more"15:54
asettleMore implies they're already contributing maintenance resources? Is that what we want to ask?15:56
asettleDo we not want to ask why they're not contributing at all?15:57
asettleor have I missed the point?15:57
zanebit implies that we're not accusing them of contribution zero, but still asks them why they're not contributing more than zero :)15:58
zanebtyping is hard15:58
asettleAh I get you15:59
*** jpich has quit IRC16:09
*** dtantsur has quit IRC16:26
*** dtantsur has joined #openstack-tc16:26
*** diablo_rojo has joined #openstack-tc16:31
*** dtantsur has quit IRC16:32
*** jamesmcarthur has quit IRC16:34
*** dtantsur has joined #openstack-tc16:37
*** ricolin has quit IRC16:45
*** irclogbot_1 has quit IRC16:46
*** irclogbot_3 has joined #openstack-tc16:47
*** sarob has joined #openstack-tc17:06
*** jamesmcarthur has joined #openstack-tc18:34
*** jaypipes has quit IRC18:37
*** jamesmcarthur has quit IRC18:38
*** sarob has quit IRC18:43
*** jamesmcarthur has joined #openstack-tc18:47
openstackgerritTim Burke proposed openstack/governance master: Add liberasurecode and pyeclib as Swift team deliverables  https://review.opendev.org/65715419:04
*** jamesmcarthur has quit IRC19:17
*** jamesmcarthur has joined #openstack-tc19:31
*** ijolliffe has quit IRC19:36
*** jamesmcarthur has quit IRC19:39
*** jamesmcarthur has joined #openstack-tc19:50
*** e0ne has joined #openstack-tc20:09
*** e0ne has quit IRC20:10
*** jamesmcarthur has quit IRC20:23
*** jamesmcarthur has joined #openstack-tc20:54
*** logan- has quit IRC21:00
*** logan- has joined #openstack-tc21:03
*** jamesmcarthur has quit IRC21:05
*** jamesmcarthur has joined #openstack-tc21:17
*** sarob has joined #openstack-tc21:20
*** jamesmcarthur has quit IRC21:21
*** lbragstad has quit IRC21:45
*** diablo_rojo has quit IRC21:54
*** sarob has quit IRC22:06
*** sarob has joined #openstack-tc22:19
*** serverascode has quit IRC22:35
*** serverascode has joined #openstack-tc22:36
*** samueldmq has quit IRC23:31
*** whoami-rajat has quit IRC23:34
*** jamesmcarthur has joined #openstack-tc23:36
*** sarob has quit IRC23:36

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