Thursday, 2018-07-12

*** edmondsw has joined #openstack-tc00:43
*** edmondsw has quit IRC00:48
Jeffrey4lTheJulia, hi i'm back, call me when you are free01:40
*** mriedem is now known as mriedem_away01:45
*** ricolin has joined #openstack-tc02:11
*** edmondsw has joined #openstack-tc02:32
*** edmondsw has quit IRC02:36
*** spsurya_ has joined #openstack-tc03:32
*** edmondsw has joined #openstack-tc04:20
*** edmondsw has quit IRC04:24
*** e0ne has joined #openstack-tc05:16
*** e0ne has quit IRC05:18
*** ricolin has quit IRC05:52
*** ricolin has joined #openstack-tc05:52
*** e0ne has joined #openstack-tc07:27
*** jpich has joined #openstack-tc07:42
*** tosky has joined #openstack-tc08:03
*** mriedem_away has quit IRC08:20
*** cdent has joined #openstack-tc08:26
mnaserAn interesting read08:31
* cdent adds to reading list08:58
cmurphyinteresting document coming from the kernel community08:58
cmurphy"Write insulting posts with a sneer just to get into a good flame every once in a while, and you’ll feel cleansed." that sounds totally healthy~08:59
*** cdent has quit IRC10:54
*** jaypipes has joined #openstack-tc11:22
*** rosmaita has joined #openstack-tc11:53
*** mriedem has joined #openstack-tc12:01
*** ricolin has quit IRC12:03
TheJuliaJeffrey4l: o/12:05
*** edmondsw has joined #openstack-tc12:09
*** dklyle has quit IRC13:14
dimsoh boy (kernel manager doc)13:47
Jeffrey4lTheJulia, o/14:09
TheJuliao/ Jeffrey4l14:14
TheJuliaI see you replied to my email, but I've not had a chance to get back to my email since noticing that I had a reply :\  It has been a busy week14:14
Jeffrey4l TheJulia got minutes? i am ok.14:14
TheJuliaI have time!14:14
TheJuliaSo as my email indicated, we're looking at the health of the projects and trying to get a feeling for current issues/challenges that the projects are facing, in order to identify if there is something we can assist or help with. I was wondering what your perception is for the current health of kolla?14:16
Jeffrey4lTheJulia, i though kolla is health as before.14:17
Jeffrey4lthere are lots newbie in kolla project14:17
Jeffrey4land the number of commits and review seems not bad.14:18
Jeffrey4lAnd we have implement several important feature during this cycle.14:19
TheJuliado you feel there is anything that is hindering or making it difficult for the kolla team?14:20
Jeffrey4lTheJulia, i think guess no so far :)14:22
dimsJeffrey4l : TheJulia : hi14:24
TheJuliaJeffrey4l: That is great to hear :)14:24
TheJuliaGood morning dims!14:24
Jeffrey4l hi dims14:24
dimsJeffrey4l : have folks moved on from kolla-kubernetes?14:24
dimsnothing left to do there i hope14:25
Jeffrey4ldims, kolla-k8s have been killed. ;( , at that time, only rich is working on it. and i guess him already moved to openstack-helm now.14:25
dims++ Jeffrey4l14:26
dimsJeffrey4l : are there any new projects showing interest on depending directly on kolla/kolla-ansible14:30
Jeffrey4ldims, kayobe ( not sure whether you know this project )14:31
dimsdon't think i have heard of it Jeffrey4l ... looking14:31
*** ricolin has joined #openstack-tc14:32
Jeffrey4lDeployment of containerised OpenStack to bare metal using kolla and bifrost14:32
dimsgotcha thanks Jeffrey4l14:34
*** mugsie has quit IRC14:36
*** mugsie has joined #openstack-tc14:36
*** mugsie has quit IRC14:36
*** mugsie has joined #openstack-tc14:36
dimsJeffrey4l : looks like kolla is coasting along, that's good news :) hopefully we can find more interest and contributors from kayobe and other projects. will there be a presence for kolla at PTG?14:37
*** hongbin has joined #openstack-tc14:37
Jeffrey4ldims, yes kolla will be there ;D14:38
dimsperfect, let's talk more when we are there14:38
Jeffrey4ldims, sure14:38
*** AlanClark has joined #openstack-tc14:58
dhellmanntc-members: who's around for office hours?14:59
*** e0ne has quit IRC14:59
TheJuliao/ although I need coffeeeeeee14:59
*** jeremyfreudberg has joined #openstack-tc15:01
dhellmannwhat's on everyone's mind this week?15:01
TheJuliaI was wondering if you had any thoughts to the question I raised ?tuesday? night15:02
AlanClarkAlanClark is listening in15:02
dhellmannhi, AlanClark15:02
TheJuliaI've not had a chance to reply to the ptl that inquired yet, and some of the discussion wednesday morning seemed interesting at least15:03
TheJuliao/ AlanClark15:03
dhellmannTheJulia : I guess I missed you question, do you want to repeat it?15:03
TheJulialet me find the log entry because it also expressed my perception15:04
TheJuliabtw, should we start the meeting for the office ours?15:04
* dhellmann is also scanning the tuesday logs
*** jeremyfreudberg| has joined #openstack-tc15:04
dhellmannI thought we decided we didn't want to use meetbot any more15:04
TheJuliadhellmann: oh, okay15:05
dhellmannbecause that hurt the casualness15:05
*** jeremyfreudberg has left #openstack-tc15:05
TheJuliaI missed that somewhere along the way15:05
TheJuliaLikely when I was taking some PTO15:05
*** jeremyfreudberg| is now known as jeremyfreud|web15:05
dhellmannwell, in on of my summaries I suggested that we just keep doing it but then in office hour logs it looked like others didn't want to15:06
dhellmannand we did discuss it in channel, though I don't remember if that was a week ago15:06
TheJuliaI concur it does hurt casuality15:06
* TheJulia goes and makes coffee15:06
dhellmannbasically I was only using the bot to get convenient log URLs and I can just paste the log file links for that15:06
dhellmannah, so the question about how to raise visibility of projects15:07
mnaserInterestingly enough on that subject, giving my very last minute prepared talk,  a lot of people assume OpenStack == VMs15:08
dhellmannare there specific projects we think need more visibility?15:09
mnaserI wouldn’t want to point out specific projects15:10
dimstangentially related, starting-yet-another-integration-point-for-k8s-and-openstack-work - - this one is to deploy a cluster by talking to nova etc (and then start k8s on the cluster just deployed)15:10
dhellmanndims : how is that different from magnum? because it starts with the kubernetes API endpoint?15:10
mnaserBut in general pushing for more presence of other projects beyond the main big ones15:10
dhellmannmnaser : presence where? in events? or deployments?15:11
mnaserdhellmann: events, marketing speak from the foundation, superuser articles15:12
dimsdhellmann : the k8s community is standardizing around this new CRD pattern for starting/maintaining clusters. eventually magnum may adopt it as well.15:12
mnaserAs much as I’d love to give more exposure to these projects, I don’t think we have the resources or ability to do much as TC other than ask nicely.15:12
dhellmanndims : nice15:12
dhellmannI guess I'm not sure of the source of concern with visibility. The foundation staff seems to go out of their way to give project teams opportunities to speak (project updates) and write for the blog. Am I mistaken? Are there folks who want to do one of those things but don't know how?15:13
dimsmnaser : we put people in touch with each other well (case in point openlab bootstrap). so if we do want to evangelize, we need to figure out what to evangelize15:14
TheJuliadhellmann: I feel like there is a huge barrier to gaining knowledge of the existence to projects from the outside of the community, and I guess I feel like providing more use case visibility might be more beneficial to projects because there are multiple was things can be bolted together15:17
TheJuliadhellmann: I'm not sure project updates are an avenue to catch attention of someone randomly trying to solve their problem of the week15:17
dhellmannour effort to produce a set of "constellations" has stalled out, would reviving that help?15:18
dhellmannTheJulia : sure, and the project updates I watched were all *very* boring, so I'm not sure how much anyone really watches those anyway15:18
TheJuliaI wonder if that might be too meta...15:18
dimsseparate identities for projects TheJulia? (their own in addition to their logo and other things)15:18
TheJuliadims: that is an interesting concept15:19
dhellmanndims : can you elaborate on why that might help?15:19
* dhellmann feels like he's missing a lot of context for this conversation15:19
dimsfolks tend to lump everything into one "openstack" bucket15:20
dimsmay help highlight strengths of individual projects15:20
dimsoutside of openstack dev community15:20
dhellmannhow would anyone find if they aren't finding
dhellmannwhat else would we need to do differently, I guess15:21
dhellmannI don't think it's a bad idea, I just don't think I fully understand it15:22
TheJuliaproject.o.o could just redirect to although they wouldn't have different indexing in that case15:22
dimsdhellmann : it's more like the style of ASF and CNCF each project has a distinct identity15:22
dhellmannis that what we want?15:23
TheJuliaI think the key is to highlight strengths and use cases of each project, and make that easy to reach... so maybe in ironic's case it would be replace our main doc landing page with something WAY less dense and that doesn't dive into every detail15:23
dimsdunno yet. thinking aloud if that would help15:23
dhellmannwe've recently had conversations about ways to break down walls between project teams to accomplish work across the community, so I'm interested to understand how separate websites for each project intersects with that goal15:24
TheJuliadoing something like that would lower the "omg that is a novel on a page" shock that may be pushing people away15:24
dhellmannTheJulia : restructuring the docs landing pages seems like it would be healthy. Did you look at Petr's proposed template?15:24
dhellmannlet me find that link15:25
TheJuliaI suspect it does not, but I also suspect people will only work across the community as it relates to their day job or interests15:25
TheJuliadhellmann: what came out of atlanta?15:25
TheJuliaatlanta ptg15:25
dhellmannwell, I think the point of that cross-community idea is that we've reached a point where implementing a feature requires work in more than one project at a time, so focusing on features rather than projects would let us pull together dynamic teams15:26
dhellmannnot that we want everyone to start working on all projects15:26
dimsgood point15:26
TheJuliathat still looks like a wall of information that context is needed to be able to parse :\15:26
dhellmannI'm sure pkovar would like to have that feedback15:26
TheJuliaoh, the template15:27
dhellmannI guess the question is what purpose the landing page on docs.o.o is meant to achieve15:27
scasi've been struggling to deal with documenting my little corner of chef, and how that should go on docs.o.o -- this seems relevant to my interests15:27
TheJuliaThe discussion I was in in atl was very much along those lines, streamlined minimized link counts, let people explore instead of trying to present it all. Sadly I've gotten resistance to simplicity. :\15:28
dhellmannscas : as part of the doc migration work last year we came up with a few audience-based divisions for the documentation; have you looked at that work?15:28
scasdhellmann: i've done some of the division in my documentation, having used other deploy space projects as inspiration15:29
dhellmannTheJulia : maybe another project team would be more receptive to going first, and then the ironic team would have an example to look at?15:30
TheJuliaMaybe part of it is in a sense, a project needs to be able to "sell" its value, if that makes sense15:30
dhellmannscas : ah, consistency can be good15:30
scasat the same time, how i document the past 2+ years of work that i've been involved in, that's something else entirely. i've taken to having a point of entry as a 'quickstart' document, a tl;dr on how to get to 'openstack' if you will15:30
TheJuliadhellmann: That might work, or I might be able to build consensus now, just haven't tried recently15:30
dhellmannTheJulia : marketing can be hard, but I think you're right it would give us some value15:31
dhellmannscas : that sounds like the ideal sort of documentation for a deployment project15:31
TheJuliaIf main pages were refined with more problem/solution oriented strings, that would also likely change some of the results people were encountering searching for solutions to their higher level problems. Then it would be key for projects to then convey that contirbuting is easy, that you don't need to be a hardcore code developer, etc15:32
dhellmannspeaking of docs, did anyone have any questions about my PTI change proposal?
TheJuliaconsidering that bit ironic, I'm all for going back to using tox15:33
dhellmannyeah, we definitely recognize that we're reverting a mistake. one difference is the specification to have a "docs" environment, instead of using venv15:33
dhellmannI consider that PTI change a pre-req to the python3-first goal, so I would like to approve it before the PTG15:34
TheJuliaNow, getting ironic back to using the warning is error flag... that is another issue entirely :\15:34
TheJuliai think it should be easy to pass the pti revision before the ptg15:35
dhellmannmost of the questions seem to be a form of "didn't we just finish going the other way?" so as long as we can overcome that sentiment I think we'll be OK15:36
dhellmannmany projects have already landed the tox.ini changes they will need15:36
* EmilienM jumps late15:38
dhellmannI think needs 1 more code review, too. It would be good if someone would doublecheck my logic against the charter rules15:38
dhellmanno/ EmilienM15:38
scasi think chef is even set up for that, without having anything published15:38
pabelanger+1 on 58049515:38
dhellmannscas : for the docs PTI change?15:39
pabelangerI believe we need to keep an eye out for projects adding commands, but not a blocker15:39
dhellmannyeah, I'm not sure how concerned I am about having teams extend their doc build process15:39
dhellmannadding doc8 for example wouldn't bother me15:39
dhellmannthat was a much bigger concern for package builds, frankly15:40
*** e0ne has joined #openstack-tc15:40
scasdhellmann: i believe so. i'm mostly eye-comparing to the proposed change and my nascent docs15:40
dhellmannscas : if you can run "tox -e docs" and get content in docs/build/html you should be all set15:41
scasonce i get tox again (new machine)15:41
pabelangerdhellmann: mostly raise it because it was an issue at some point as I understand correctly15:41
pabelangerI think fungi has more history15:41
dhellmannI saw tonyb started setting up the election repo for upcoming elections,
dhellmannpabelanger : it was more of an issue when the CI system didn't use tox, but yeah15:42
dhellmannthe bigger issue is actually the hackery involved in installing some projects' dependencies15:42
dhellmannand this should help us get back to a state where we can do that correctly everywhere15:42
scasdhellmann: yep. 'tox -e docs' works15:43
dhellmannanother thing we should talk about on the schedule is when we want to make the final call on goals for stein15:43
pabelangeryes, agree. It also should help remove some of the legacy zuul-cloner some jobs are still using15:43
dhellmannsmcginnis_afk is taking the lead on that, is someone helping him?15:43
dhellmannscas : \o/15:43
dhellmannpabelanger : yep15:43
fungiabout to start another meeting, but catching up now15:44
jrollTheJulia: I hate that we compare everything to k8s, but something more like is what you're thinking?15:44
dhellmannTheJulia : I'm reviewing the tracker, and I see the anti-nitpicking campaign on that list still. Do you consider that "done" for now, or do you have more work planned (so that I should keep it on the tracking list)?15:44
jrollTheJulia: like more.... "marketing" focused web sites for projects15:45
dhellmannlike ?15:45
TheJuliajroll: not necessarilly marketing, but less information overload15:45
jrollTheJulia: yeah, hence the air quotes... more like "ironic solves x, y, and z. link to getting started, link to docs, link to demo video15:46
pabelangerdhellmann: I also believe I've carved out a timeslot to help monitor more on the python3 changes for zuul, so if you'd like to increase your ping window, I'll be more responsive.15:46
TheJuliadhellmann: I think there was another thing with that I wanted to address, I'll have to find my notes, but if you don't hear from me before monday saying oherwise, it can be considered done15:46
dhellmannpabelanger : that's good news, thanks!15:46
TheJuliajroll: kind of I guess15:47
dhellmannTheJulia : how about if you update the wiki with a note about what else you want to do, or to remove the block?15:47
TheJuliadhellmann: that works!15:47
jrollTheJulia: ok, I guess I'll have to see your ideas on a whiteboard or something :)15:47
TheJuliajroll: heh, okay15:48
* dims back reading backscroll15:48
fungithe original concern with tox -e docs is that some projects had started to add random extra scripted steps to build their documentation for things they could have done as sphinx plug-ins but adding random shell commands and assumptions was "easier"15:48
fungiis that the history you were asking about, pabelanger?15:49
pabelangerfungi: yes, but moving back to tox -edocs, that is still a concern or not any more15:49
*** zaneb has quit IRC15:49
dhellmannwas that a problem because those steps were in the "docs" env but we built docs using the venv env?15:50
*** zaneb has joined #openstack-tc15:50
fungiit's still a concern, but we just need to be more vigilant in reminding people it's a bad idea to cram extra arbitrary build steps in their docs env15:50
fungias far as the idea, we _used_ to have those for all projects many years ago because the docs team only wanted manuals hosted on docs.o.o and project-controlled documentation had to go on separate sites. eventually we came to an agreement that we could publish project docs to the main docs site and set up permanent redirects from projectname.o.o to docs.o.o/developer/projectname15:51
fungiand eventually were able to drop those redirects once they were no longer getting used15:52
TheJuliafungi: ahh15:52
TheJuliagood to know15:52
fungiit took many years to finally get it all consolidated15:52
fungiand cleaned up15:53
scasfrom a deploy project perspective, i'd prefer a breadcrumb from docs.o.o and not my own custom landing page15:53
TheJuliaall consolidated likely helps the algorithms upvote the information because it falls into a single domain name15:53
dhellmannscas : is chef not listed on docs.o.o with the other projects?15:53
scasnot yet15:53
dhellmannok, let me or pkovar know if you need help with that15:54
fungiwell, also it means if you want https-hosted docs you don't need ssl certs for every project15:54
dhellmannfungi : do you happen to know if the work on the sitemap.xml was ever finished?15:54
scasi think i just need to flip the job from build to publish, but i've not had a good solid review of what i have merged to make sure it works from an actual structural/docs.o.o point of view15:54
fungiand don't need to make dns changes every time you add/remove/rename a project15:54
fungidhellmann: no clue15:55
dhellmannfungi : yes, having 1 cert seems like a good simplification in terms of admin for the infra team15:55
* fungi apologizes, intermittent connectivity and about to go into a lengthy meeting tunnel15:55
dhellmannok, I'll see if the docs team knows15:55
*** jeremyfreud|web has quit IRC15:55
dhellmannscas : the docs team may be able to give you some advice there. pkovar made an offer for that sort of thing on the -dev list a while back, irrc15:56
scasdhellmann: ok. i'll consult my archives15:56
fungifor years i've felt that even having a separate docs.o.o site is suboptimal, but it's an artifact of the separate jurisdiction with www.o.o basically being the foundation website rather than the community site15:57
dhellmannhmm, yeah15:57
fungifor zuul, docs are hosted at because the community operates that site and can publish their documentation to the same place as other project content15:57
scaswww.o.o/docs/... might go over better for seo purposes, too.15:57
dhellmannperhaps if the foundation is renamed as a part of the expansion, we can make that sort of change and take over more of the domain15:58
fungithe jurisdiction issue over www.o.o is why we have separate security.o.o and governance.o.o and releases.o.o and specs.o.o and so on rather than being able to serve them all from one vhost15:58
dhellmannalthough I'm not necessarily looking forward to reorganizing the website content again15:58
* ttx waves15:58
dhellmanno/ ttx15:59
* dhellmann needs to find some lunch16:00
scasi presume the domain is not under influence?16:01
dhellmannthat doesn't look like something we control16:01
scasyeah. it looked rather squatter-y16:01
*** dklyle has joined #openstack-tc16:02
fungisomebody in panama with hidden contact info via namecheap16:02
scasthe general theme, i guess i'm getting at, is that most other projects with a foundation attached usually have a separate domain away from project.ext for that information16:03
fungianyway, there's a lot of marketing inertia with the site at the moment, so if there ever were a foundation rename i wouldn't expect any sort of hand-off to go quickly16:03
fungibut for comparison, is the site for the asf and is the site for the apache web server16:05
fungiso i don't know that what osf has is all that different (even if i find it suboptimal from a community project perspective)16:08
*** annabelleB has joined #openstack-tc16:08
TheJuliafungi: I seem to remember -> was over the time span of years16:08
fungiyeah, that sounds familiar16:09
TheJuliaIncluding a time when I think there was a "if your looking for the webserver, click here." link.16:09
dimsthey do have a central location as well - with pointers to all the other projects and their websites16:13
scasasf does look rather shiny16:16
*** mriedem is now known as mriedem_away16:16
scasbluntly, it looks like a shinier version of how openstack has grown16:17
*** e0ne has quit IRC16:20
*** e0ne has joined #openstack-tc16:23
*** e0ne has quit IRC16:23
scasif one considers 'openstack' to just be the core services, project.o.o works and makes a lot of sense from how each project is intended to be showcased. it does fall down when factoring in the deploy projects that are so unfortunately named, like chef16:25
*** spsurya_ has quit IRC16:26
*** AlanClark has quit IRC16:26
scasif i were coming at this from a new user perspective, i'd come away with a higher confidence level just by seeing the information exposed in a manner as asf, however it's structured16:31
*** jpich has quit IRC16:39
*** e0ne has joined #openstack-tc16:41
*** e0ne has quit IRC16:43
*** ricolin has quit IRC16:44
jrolloh my, guido is stepping down, will be interesting to see the evolution of python governance:
*** zaneb has quit IRC16:57
*** zaneb has joined #openstack-tc16:59
*** ricolin has joined #openstack-tc17:15
*** annabelleB has quit IRC17:32
*** rosmaita has quit IRC17:33
*** mriedem_away is now known as mriedem17:37
TheJuliaIt shall be interesting17:42
*** rosmaita has joined #openstack-tc17:43
TheJuliaSince... I think the whole thing about living in interesting times already applies.17:43
*** ricolin has quit IRC17:51
*** harlowja has joined #openstack-tc18:24
*** e0ne has joined #openstack-tc18:42
*** e0ne_ has joined #openstack-tc18:47
*** e0ne has quit IRC18:47
*** zaneb has quit IRC18:50
*** zaneb has joined #openstack-tc18:51
*** e0ne has joined #openstack-tc18:51
*** e0ne_ has quit IRC18:52
*** e0ne has quit IRC18:54
*** e0ne_ has joined #openstack-tc18:54
*** e0ne_ has quit IRC18:56
*** zaneb has quit IRC18:57
*** zaneb has joined #openstack-tc18:58
*** e0ne has joined #openstack-tc19:12
*** e0ne has quit IRC19:23
openstackgerritDoug Hellmann proposed openstack/governance master: write up the python3-first goal
*** e0ne has joined #openstack-tc19:31
*** mriedem has quit IRC19:32
*** e0ne_ has joined #openstack-tc19:34
*** e0ne has quit IRC19:36
*** e0ne_ has quit IRC19:41
*** e0ne has joined #openstack-tc19:55
*** e0ne has quit IRC19:59
*** e0ne has joined #openstack-tc20:03
*** zaneb has quit IRC20:06
*** zaneb has joined #openstack-tc20:06
*** e0ne has quit IRC20:07
*** e0ne has joined #openstack-tc20:13
*** e0ne has quit IRC20:18
fungiit does come across at least slightly as a ragequit e-mail20:19
*** e0ne has joined #openstack-tc20:21
fungiif he'd just left out that first sentence, the entire rest of the message would have had a much different tone20:24
*** e0ne_ has joined #openstack-tc20:27
*** e0ne has quit IRC20:28
*** e0ne_ has quit IRC20:31
*** e0ne has joined #openstack-tc20:35
*** e0ne has quit IRC20:42
scasit came off to me as being somewhat crabby, somewhat exhausted20:44
scasi blame the bay area monoculture20:45
*** e0ne has joined #openstack-tc20:48
*** e0ne has quit IRC20:50
*** mriedem has joined #openstack-tc20:55
*** e0ne has joined #openstack-tc21:27
*** edmondsw has quit IRC21:34
*** e0ne has quit IRC21:39
*** edmondsw has joined #openstack-tc21:40
*** edmondsw has quit IRC21:44
*** e0ne has joined #openstack-tc21:59
*** dklyle has quit IRC22:16
*** dklyle has joined #openstack-tc22:16
zanebI can't believe there are people who don't like PEP572. Have those people ever programmed in Python?22:31
*** e0ne has quit IRC22:32
*** zaneb has quit IRC22:52
*** zaneb has joined #openstack-tc22:52
*** zaneb has quit IRC22:57
*** rosmaita has quit IRC22:59
*** tosky has quit IRC23:01
*** hongbin has quit IRC23:19

Generated by 2.15.3 by Marius Gedminas - find it at!