Tuesday, 2018-05-29

*** ricolin has joined #openstack-tc00:08
*** edmondsw has joined #openstack-tc01:02
*** edmondsw has quit IRC01:07
*** gagehugo has joined #openstack-tc01:22
*** gagehugo has quit IRC01:31
*** gagehugo has joined #openstack-tc01:36
*** pabelanger has quit IRC02:40
*** pabelanger has joined #openstack-tc02:40
*** notmyname has quit IRC02:44
*** mtreinish has quit IRC02:44
*** lxkong has quit IRC02:44
*** zaneb has quit IRC02:44
*** spsurya has quit IRC02:44
*** persia has quit IRC02:44
*** portdirect has quit IRC02:44
*** lxkong has joined #openstack-tc02:45
*** zaneb has joined #openstack-tc02:45
*** spsurya has joined #openstack-tc02:45
*** persia has joined #openstack-tc02:45
*** portdirect has joined #openstack-tc02:45
*** mtreinish has joined #openstack-tc02:45
*** mtreinish has left #openstack-tc02:45
*** mtreinish has joined #openstack-tc02:45
*** notmyname has joined #openstack-tc02:46
*** evrardjp has quit IRC02:46
*** tonyb has quit IRC02:46
*** openstackgerrit has quit IRC02:46
*** purplerbot has quit IRC02:46
*** alex_xu has quit IRC02:46
*** frickler has quit IRC02:46
*** evrardjp has joined #openstack-tc02:47
*** tonyb has joined #openstack-tc02:47
*** openstackgerrit has joined #openstack-tc02:48
*** frickler has joined #openstack-tc02:48
*** purplerbot has joined #openstack-tc02:48
*** alex_xu has joined #openstack-tc02:48
*** dtroyer has quit IRC02:50
*** mgagne has quit IRC02:50
*** fanzhang has quit IRC02:50
*** jroll has quit IRC02:50
*** dtruong has quit IRC02:50
*** gagehugo has quit IRC02:50
*** ianychoi_ has quit IRC02:50
*** jaosorior has quit IRC02:50
*** corvus has quit IRC02:50
*** aspiers has quit IRC02:50
*** mgagne has joined #openstack-tc02:50
*** dtroyer has joined #openstack-tc02:50
*** fanzhang has joined #openstack-tc02:50
*** jroll has joined #openstack-tc02:50
*** dtruong has joined #openstack-tc02:50
*** edmondsw has joined #openstack-tc02:51
*** gagehugo has joined #openstack-tc02:51
*** ianychoi_ has joined #openstack-tc02:51
*** jaosorior has joined #openstack-tc02:51
*** corvus has joined #openstack-tc02:51
*** aspiers has joined #openstack-tc02:51
*** edmondsw has quit IRC02:55
*** jaosorior has quit IRC07:48
*** dtantsur|afk is now known as dtantsur08:04
*** jaosorior has joined #openstack-tc08:32
*** ianychoi_ is now known as ianychoi08:40
cmurphyo/ hiya tc-members09:02
* ttx grabs a coffee09:04
*** ricolin has quit IRC09:04
ttxI don't have much to discuss, still processing the wek09:07
* cmurphy same09:07
cmurphyso many notes to go through09:08
openstackgerritThierry Carrez proposed openstack/governance master: Clean up remaining docs links  https://review.openstack.org/57087809:32
*** rosmaita has joined #openstack-tc12:03
*** edmondsw_ has joined #openstack-tc12:11
*** kumarmn has joined #openstack-tc12:17
*** dklyle has quit IRC12:42
*** tobberydberg has joined #openstack-tc12:47
*** kumarmn has quit IRC12:48
*** kumarmn has joined #openstack-tc12:49
*** kumarmn has quit IRC12:53
*** kumarmn has joined #openstack-tc13:06
*** mriedem has joined #openstack-tc13:23
*** dtantsur is now known as dtantsur|brb13:31
*** lbragstad has joined #openstack-tc13:32
*** edmondsw_ is now known as edmondsw13:42
mnasertc-members: forgive my small ignorance regarding licenses, but can we import/run ci on gplv2 and gplv3 projects?13:53
openstackgerritJulia Kreger proposed openstack/governance master: Add principles entry for nitpicking  https://review.openstack.org/57094013:53
mnaserthis is in regards to the starlingx discussion of dtroyer splitting the project imports and going over the review, i noticed the import of https://github.com/starlingx-staging/stx-gplv2 and https://github.com/starlingx-staging/stx-gplv313:53
fungimnaser: sure, they just can't ship as part of openstack or one of its runtime dependencies13:53
ttxmnaser: openstack rules at https://governance.openstack.org/tc/reference/licensing.html13:54
ttxother rules may vary13:54
mnaseri had that document open, i just wasn't sure on the call to make on the infra import review so figured i'd consult and ask13:54
ttxdefinitely OK for projects hosted on our infra13:54
ttxthere are a bunch of gplv2 / gplv3 things there13:55
smcginnisHosting GPL code in one of our repos scares me a little. That's different than using GPL libraries.13:55
fungiright, the licensing rules we have are mainly about what can be an official part of openstack. the infra hosting of projects is open to anyone developing under an osi-approved license13:55
fungismcginnis: you'll have to be more specific about what you mean by "our repos" in this case13:56
ttxA lot of the infra repos are GPL13:56
mnaserfwiw, i am speaking about this review -- https://review.openstack.org/#/c/569562/13:56
smcginnisfungi: Yeah, the word "our" in general has been problematic in these discussions.13:56
smcginnisI mean the chance of something seen as part of OpenStack but being GPL seems like it creates a risk that someone might copy or cherry pick something from one hosted repo to another. I'll leave it to the more lawyerly types to say if that's an issue for us or not.13:57
ttxsmcginnis: the lawyerly approved https://governance.openstack.org/tc/reference/licensing.html13:58
ttxand that's just about requirements vs. test-requirements13:59
jrolleh, people are going to copy/paste without reading licenses, whether or not it's in the openstack git infra13:59
ttxeverything we do not depend on but that is hosted in project infra is not openstack13:59
mnaseri went over the repos and i think we'd want to make sure stuff like this is not part of the import, so i'll put that in a comment -- https://github.com/starlingx-staging/stx-utils/blob/master/README.confidentiality.txt14:00
*** jroll has quit IRC14:01
*** jroll has joined #openstack-tc14:02
*** mriedem is now known as mriedem_away14:13
*** hongbin has joined #openstack-tc14:42
zanebfungi: question for you on the mailing list: http://lists.openstack.org/pipermail/openstack-dev/2018-May/130807.html14:58
*** dtantsur|brb is now known as dtantsur15:00
clarkbre GPL runtime dependencies openstack has many, they just aren't linked against15:01
fungizaneb: cool, i'll follow up to that in a bit. it's particularly complicated by the fact that we want to spin the infra hosting effort out of openstack tc governance (for what should be increasingly obvious reasons) and so need somewhere distinct to put that stuff15:01
zanebfungi: yeah, agreed. but AIUI it is our responsibility for now as the only TC for the foundation15:02
clarkbzaneb: fungi there is the old stackforge doc which I have an update to call unofficial project hosting which I'm not sure if it merged or not15:02
zanebwhen the foundation has multiple TCs it will be much more complicated15:02
fungiobviously tc resolutions won't do much good for an effort which is outside the tc's direct control. we're working on ideas now to be able to take input from the various technical committees15:02
fungizaneb: the only one which calls itself a technical committee maybe, but that's just semantics15:03
fungithe kata project has a steering committee, for example15:03
fungithe openstack tc doesn't really hold sway over kata or zuul15:03
fungiother than, you know, because there is a good reason to collaborate and take the openstack community's needs into account15:04
zanebfungi: the bylaws define the board's relationship with the TC, which currently means us. agree that we have no jurisdiction over projects adopted directly by the board into the new focus areas15:05
zanebthe question here is how to handle projects that aren't official in *any* of the focus areas, including OpenStack15:06
fungiright, that's basically what this is. for the infra team efforts, we want a way for each of those (including openstack) to have input, but as such can't really be directly governed by any of them15:06
zanebyeah, that will be a necessary step15:07
fungiso, as i said, all good things we should document, but documenting them under the openstack tc is not the right way to go about it looking forward15:07
zanebfungi: I agree with that for the near future, but it's the only way for Right Now, isn't it?15:09
smcginnisTC really has nothing to do with these other project.15:09
fungithe infra team is working to get all implication of openstackness removed from the services its hosting so that the openstack community is merely another tenant of the hosting infrastructure, but it takes time to do things like pick a domain name and move services in non-impacting ways15:09
smcginnisSo most of this is just us talking as community members of the foundation.15:09
clarkbhttps://docs.openstack.org/infra/system-config/unofficial_project_hosting.html that doc did lose its "stackforge" strings15:09
smcginniszaneb: We don't actually have any say on things like starlingx other than being "concerned citizens".15:10
zaneblet's leave starlingx out of the discussion :)15:10
fungiwell, and voting for or against board members who agreed on things which make starlingx possible15:10
clarkbfwiw starlingx seems to fit right into the purpose of unofficial project hosting as listed on that doc15:10
clarkbbasically to foster collaboration between what is and isn't openstack15:11
smcginniszaneb: It's part of this though, because unless something is in https://github.com/openstack/governance/blob/master/reference/projects.yaml - the TC has no real involvement.15:11
clarkbthrough the use of a common set of tools that (we feel) are awesome and useful15:11
zanebsmcginnis: the foundation is ultimately responsible for what is hosted on foundation infrastructure, and the for now the Foundation takes its technical advice from the OpenStack TC. Doesn't it?15:12
fungizaneb: so anyway, i'll follow up to that ml thread with a proposal that we prominently link our hosting requirements and update https://docs.openstack.org/infra/system-config/unofficial_project_hosting.html as needed to reflect what it might be missing15:12
smcginnisThe Foundation does not take technical advice from the TC, only looks to the TC for technical leadership of OpenStack-proper.15:12
fungiit's fair to note that the foundation bylaws will need updating to reflect that there are other projects besides openstack managed by the osf15:13
fungia lot of that assumption is baked into the bylaws currently15:13
fungiso the openstack tc features prominently in the bylaws due to that assumption15:14
*** dklyle has joined #openstack-tc15:15
fungiwhen they first formed the foundation they didn't consider that there might be other efforts outside openstack worth caring about. heck, when i started on the project the client libs, qa and ci tooling, et cetera weren't considered official projects and the board and tc had to agree that it was okay that we were hosting their git repositories15:16
zanebsmcginnis: _technically_ infra is an official part of OpenStack-proper, soooo... ;)15:17
zanebfor now, at least15:17
smcginnisNo, not really.15:17
fungiand figuring out where the fuzzy edges were begat the idea that we'd have a "stackforge" to put unofficial stuff in15:17
smcginnisIt's part of the foundation.15:17
zanebprojects.yaml disagrees15:17
fungiwell, officially for now the infrastructure team is an official team under the tc15:17
fungizaneb is correct in that regard15:18
cmurphyclarkb: I think starlingx isn't just looking for hosting, they're also intending to be a part of the foundation under the edge strategic focus area, so not exactly unofficial iiuc15:18
fungiit's part of openstack technical governance, but not part of openstack the shipping product (if there even is one)15:18
fungicmurphy: well, "unofficial" from the historic perspective of the openstack technical committee15:19
fungione person's unofficial is another person's official15:19
zanebcmurphy: regardless of what they _want_ to happen, it's completely unofficial until the foundation staff agrees to accept them into that focus area, which AIUI has not happened, although it's super unclear even to me, let alone the rest of the world15:20
fungibut yes, it all drives back to the currently uneasy position with suggesting to these other non-openstack projects that it's cool to host your development on the openstack community infrastructure, which is under the direct control of the openstack technical committee15:20
fungiand to cmurphy's point, we probably need to start saying things are "not openstack" rather than "unofficial" (or be clear that we mean "unofficial from the perspective of what the openstack technical committee has declared as official openstack projects)15:21
zanebfungi: so here's my point: it is completely unfair to the infra team to force them to make all the decisions about which projects the foundation should be hosting, without any guidance15:24
cmurphyI guess in my head I still see an invisible stackforge namespace where it really is okay to just host things without any official blessing and definitely without marketing assistance, and that space is universally "unofficial"15:24
zanebbecause if e.g. some sort of massive PR disaster results, responsibility should be taken by a group that is accountable for not causing such disasters (i.e. us) and not by Infra15:25
zanebcmurphy: I totally agree, and we need to make sure the projects that are in that space see it that way too :)15:26
fungizaneb: yep, all good reasons for us to get the name "openstack" removed from all our infrastructure as soon as we can15:26
fungicmurphy: to me, unofficialness is ephemeral. projects go from being unofficial to official or back over time anyway15:27
fungithey shouldn't be forced to significantly change how their development is hosted when that happens15:28
fungi(noting, for example, that kata is "official" in the sense that it is now a container strategic focus area pilot project managed by the osf, but still does its development on github)15:29
clarkbzaneb: yes, I think guidance there can help. That said I do somewhat feel like the guidance we currently have is largely sufficient for the current problem space. The problem seems to be that people disagree on the guidance allowing us to do the hosting for eg starlingx15:33
zanebclarkb: what is the current guidance as you understand it?15:34
clarkbzaneb: osi approved license, intent to collaborate with openstack "proper"15:34
clarkbif I had to boil it down to a one liner15:34
clarkb(granted mnaser pointed out where they may have to clarify the licensing bits but I expect that they will do that as they've explicitly said it will be open source)15:35
clarkbwith my Infra hat on the confusion is popping up because people seem to disagree with the existing guidance15:37
clarkbfor example there was specific confusion over using the "openstack/" namespace rather than "stackforge/"15:37
zanebok, so it seems we're holding starlingx to a higher bar than that, including asking them to contact all the upstreams (not just OpenStack) that they have forked or carry patches for15:38
zaneband I think that's partly because we haven't encountered this situation before15:38
clarkbzaneb: yup, I do think this particular situation does present some unique challenges. And I think it is also ok to say existing guidance isn't strong neough and work from there15:39
zanebbut it seems like a good opportunity to flesh out the guidance, and give infra a good starting point to build from once they go out on their own (outside of the OpenStack TC's purview)15:39
zaneband the other part is, we are offering something valuable to these projects, it's totally OK to ask for something in return even if they are not bound by our governance. explicitly agreeing to be bound by the Foundation CoC and also not using the OpenStack trademark in potentially misleading ways are very reasonable asks IMO15:41
zanebalso I apologize to everyone for using 'ask' as a noun just then15:42
fungii agree that we're holding starlingx to a higher bar, but in this case it's that the openstack tc is providing recommendations to the osf related to fostering collaboration with the openstack community, relevant in particular because the software in question integrates with (a forked version of) openstack15:47
fungiso these aren't necessarily requirements for being hosted, but rather guidance on how best to go about it in this particular case (with the osf possibly passing those requirements on to the starlingx team as conditions for being managed by the osf)15:49
zanebfungi: on Friday we also asked them to contact all the other upstreams they had forks of/patches to as well, because we don't want the foundation to appear to be going around forking a bunch of projects without talking to them. so it can't be just that we're looking after our OpenStack turf only15:51
zanebfungi: actually, are you saying that Infra received a request from the Foundation directly to host StarlingX?15:52
fungizaneb: nope, the infra team received a request from dtroyer to host some repositories15:53
zanebok, cool, that's what I originally thought15:53
fungiyou and i and lots of other people then ended up in a room at the very end of the forum providing our recommendations to the osf for what it should be requiring of the starlingx team15:54
fungieasy to conflate requirements for hosting some repos with requirements the osf might impose on a new project requesting to be managed (which might include telling it how to go about hosting its repos)15:55
zanebone could certainly read the press release as claiming that the board itself had agreed to the hosting, and I'm sure many people will :/15:55
fungiyeah, i had talked to the foundation executives and that wasn't going to be a press release from the openstack foundation. they were going to make a big splash about kata and zuul, but airship and starlingx were simply mentioned at the board meeting (partly at my urging) to get in front of the inevitable surprise many board/tc/uc members and others who attend board meetings15:57
fungiwere sure to experience15:57
fungithen intel decided to stick it into their keynote at the last moment15:58
fungiwhich didn't help matters, but was thankfully only in passing15:58
*** mriedem_away is now known as mriedem16:11
*** mriedem has quit IRC16:15
*** mriedem has joined #openstack-tc16:20
fungimainly i wanted to be sure we could have these exact sorts of conversations without it looking like a fait accompli16:25
dimsthanks for that fungi16:25
*** eandersson has quit IRC16:53
*** dklyle has quit IRC17:12
fungiunfortunately marketing and reality often suffer a bit of an impedance mismatch17:40
*** jaosorior has quit IRC17:49
dhellmannmnaser, EmilienM_PTO : please go ahead and write up that email, since you'll be monitoring the work on starlingx17:59
dhellmannfungi , zaneb , clarkb : I mostly agree with fungi's summary of the situation. I would like us to make sure we have some sort of policy written down, but I'm not sure we're far enough into the current cases to extrapolate what that policy should say, yet.18:06
fungiand there's also a distinction which needs to be drawn between the tc's preference/guidance to the foundation over such matters, and the extent of our ability to legislate such things18:08
dhellmannyes, that's something for the board to clarify18:09
fungiwe have a large sphere of influence to sway the decisions taken by the foundation board, the foundation staff, and other people in our ecosystem. it's great that we leverage that to fight for a positive outcome18:09
fungibut we're not well-placed to necessarily demand such things18:10
fungii was pleased though that representatives of the board, foundation executives and starlingx development team seemed to take our feedback seriously18:11
fungithat's not to say that i was surprised they did so, which of course pleases me even more that i've come to expect such fidelity18:12
zanebto clarify, I'm not interested in grabbing power off of anyone. I don't think the board *wants* a line-item veto on which git repos infra does or does not create, and as I said earlier I also don't think it's fair to leave infra holding the bag either18:13
*** dtantsur is now known as dtantsur|afk18:28
*** lbragstad has quit IRC18:39
*** dtruong_ has joined #openstack-tc18:43
*** dtruong has quit IRC18:53
fungisure, fwiw i don't think the infra team reviewers want to be spending their time policing some complex set of rules either, so keeping things simple is probably for the best18:57
zanebyeah, +1 for as simple as possible18:59
*** lbragstad has joined #openstack-tc19:00
zanebwe don't even necessarily need to be monitoring all of the stuff, more to make it clear what projects are agreeing to (e.g. CoC) so that if we have to tell them later to go find another place to host, we're not seen as being capricious19:02
*** Guest32198 is now known as melwitt19:06
fungisure, as long as it says somewhere prominent that by hosting here you're implicitly agreeing to the code of conduct and to produce your work under an osi-approved license, that's probably enough. the infra team already reserves the right to deny service to anyone abusing our systems or otherwise negatively disrupting our community19:16
fungiwe'd rather just deal with the handful of problematic outliers than add a bunch of rules for every possible way someone might try to cause a problem19:17
openstackgerritSean McGinnis proposed openstack/project-team-guide master: Add note about checking README in pep8 tests  https://review.openstack.org/57100719:29
zanebsmcginnis: would the PTI be somewhere we should mention ^that?19:48
openstackgerritDoug Hellmann proposed openstack/governance master: add python 2 deprecation timeline  https://review.openstack.org/57101119:50
openstackgerritDoug Hellmann proposed openstack/governance master: add python 2 deprecation timeline  https://review.openstack.org/57101119:52
smcginniszaneb: Maybe. Not sure where the best place would be.19:57
smcginnisI think the pti is for "OpenStack" projects, so actually maybe somewhere under the governance repo for now?19:57
*** harlowja has joined #openstack-tc20:00
openstackgerritDoug Hellmann proposed openstack/governance master: add python 2 deprecation timeline  https://review.openstack.org/57101120:22
*** edmondsw has quit IRC21:13
dimshope we don't encourage the typo-fixers to start nit-picking (and hence editing) everyone's reviews ...21:48
openstackgerritMerged openstack/project-team-guide master: Add note about checking README in pep8 tests  https://review.openstack.org/57100721:50
dhellmanndims : that's one reason to encourage follow-up patches, I guess21:51
dims++ dhellmann21:55
fungito reiterate my earlier question from friday, if we start using meetbot for recording minutes of office hours, should we reuse #startmeeting tc for that or pick a different name like tc-office-hours?22:09
fungifigured i would have a go at it when 01:00z rolls around22:10
mnasertc-members: https://etherpad.openstack.org/p/9E4lvFpgWh -- this is a rough draft in regards to updating what's going on with starlingx to openstack-dev, the discussion around the direction to take as well as some takeaways, i'd really appreciate a read over before i send it through (dhellmann was taking some notes so appreciate your input too)22:10
fungimnaser: i wonder if we should try to stick to the approved board terminology for this and call it a "proposed pilot project for the edge computing strategic focus area"?22:15
fungihttps://wiki.openstack.org/wiki/Governance/Foundation/5Nov2017BoardMinutes has "RESOLVED, the Board authorizes the Foundation staff to incubate strategic focus areas, including pilot projects, adjacent to the OpenStack Project; provided that the strategic focus areas shall be approved by the Board."22:15
mnaserfungi: i think that would be good, i kinda wanted to go that way but i wasn't sure for the exact terminology22:15
fungimy worry is if we use different words for the same concept every time, we seed more confusion about what's going on22:16
mnaseri agree, i just didn't have the right words22:17
mnaserfungi: do you mind fitting that in where you think it ideally goes in the etherpad? :>22:17
fungisure, just didn't want to mar your pretty text without discussing first! ;)22:19
*** kumarmn has quit IRC22:22
fungialso noted that starlingx includes forks of openstack and other open source software, it's not strictly forks of other software though22:22
fungiwhich is what the prior phrasing implied22:22
*** kumarmn has joined #openstack-tc22:23
*** kumarmn has quit IRC22:27
*** kumarmn has joined #openstack-tc22:41
*** kumarmn has quit IRC22:46
fungimnaser: lgtm, i poked a handful of other edits in there for clarity22:59
*** hongbin has quit IRC23:00
fungisince nobody has objected to reusing #startmeeting tc for office hour minutes, i'll give that a go in a couple hours23:00
smcginnisfungi: That's what's listed here: http://eavesdrop.openstack.org/#Technical_Committee_Office_hours23:00
fungioh, excellent point!23:01
fungithe decision was already made (implicitly via laziness) and i didn't even think to look there23:01
smcginnisI will have to miss this office hour since my daughter has an end of year choir concert tonight.23:02
smcginnisLaziness for the win!23:02
fungiooh, i hope she hits every note!23:02
smcginnisI chaperoned a trip with these kids to Chicago ~month ago. I'm pretty impressed with all of them.23:03
smcginnisMuch better than I could ever sing.23:03
fungii honestly don't have a topic planned and expect to be one of the only (if not the only) participants anyway23:03
mnaserI’ll stick around if a discussion comes up :)23:03
mnaserIt’s not so bad of a time for me.23:04
smcginnisIt's on my todo to summarize the Stein goal selection discussion from the Forum and get the conversation going. That should be good for one or two office hours for the near future.23:04
smcginnisAnd with that, better get going...23:04
fungiyeah, i still have a couple of session summaries i need to write before i get back to more normal work. sort of a slow-start week23:05
clarkbthe goal session was one I couldn't make. Would be interested to hear how that went when people are able to talk about it23:05
fungiit was mostly a non-deciding session... we spent a lot of time on the meta discussion of the goals process itself23:05
*** mriedem has quit IRC23:06
fungididn't get to pitching actual stein goal selection until the last 10-15 minutes23:06
*** kumarmn has joined #openstack-tc23:36
*** openstackstatus has joined #openstack-tc23:44
*** ChanServ sets mode: +v openstackstatus23:44
*** kumarmn has quit IRC23:46
*** kumarmn has joined #openstack-tc23:48
*** kumarmn has quit IRC23:49
*** kumarmn has joined #openstack-tc23:51
*** kumarmn has quit IRC23:54
*** kumarmn has joined #openstack-tc23:57

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