| *** ijolliffe has quit IRC | 00:06 | |
| *** ricolin has joined #openstack-tc | 02:39 | |
| *** tetsuro has quit IRC | 03:16 | |
| *** tetsuro has joined #openstack-tc | 03:54 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #openstack-tc | 04:33 | |
| *** jaosorior has quit IRC | 05:21 | |
| *** jaosorior has joined #openstack-tc | 05:23 | |
| *** tetsuro_ has joined #openstack-tc | 05:45 | |
| *** tetsuro__ has joined #openstack-tc | 05:47 | |
| *** tetsuro has quit IRC | 05:47 | |
| *** tetsuro_ has quit IRC | 05:50 | |
| *** dklyle has quit IRC | 06:04 | |
| *** belmoreira has joined #openstack-tc | 06:38 | |
| *** rpittau|afk is now known as rpittau | 07:18 | |
| *** ralonsoh has joined #openstack-tc | 07:21 | |
| *** tetsuro__ has quit IRC | 07:37 | |
| *** tosky has joined #openstack-tc | 07:38 | |
| *** tetsuro has joined #openstack-tc | 07:40 | |
| *** slaweq has quit IRC | 08:15 | |
| *** slaweq has joined #openstack-tc | 08:20 | |
| *** e0ne has joined #openstack-tc | 08:20 | |
| *** njohnston|pto has quit IRC | 08:52 | |
| *** tetsuro has quit IRC | 08:57 | |
| *** tetsuro has joined #openstack-tc | 09:00 | |
| *** rpittau is now known as rpittau|bbl | 09:15 | |
| *** slaweq has quit IRC | 09:19 | |
| *** e0ne_ has joined #openstack-tc | 09:20 | |
| *** e0ne has quit IRC | 09:20 | |
| *** e0ne has joined #openstack-tc | 09:28 | |
| *** tetsuro has quit IRC | 09:28 | |
| *** e0ne_ has quit IRC | 09:28 | |
| *** slaweq has joined #openstack-tc | 09:57 | |
| openstackgerrit | Merged openstack/project-team-guide master: Add deprecation of a repo https://review.opendev.org/734877 | 10:05 |
|---|---|---|
| *** slaweq has quit IRC | 10:20 | |
| *** slaweq has joined #openstack-tc | 10:26 | |
| *** slaweq has quit IRC | 10:30 | |
| ttx | gmann: I posted a PS3 on https://review.opendev.org/#/c/734074/ and you commented on PS2 -- please see latest and let me know if it addresses your concern | 10:32 |
| *** ricolin has quit IRC | 10:47 | |
| *** purplerbot has joined #openstack-tc | 10:58 | |
| *** slaweq has joined #openstack-tc | 11:24 | |
| *** rpittau|bbl is now known as rpittau | 11:54 | |
| *** slaweq has quit IRC | 12:13 | |
| *** slaweq has joined #openstack-tc | 12:13 | |
| *** ijolliffe has joined #openstack-tc | 12:17 | |
| *** slaweq has quit IRC | 12:18 | |
| *** tkajinam has quit IRC | 12:25 | |
| *** mnaser has quit IRC | 12:32 | |
| *** jungleboyj has quit IRC | 12:33 | |
| *** gmann has quit IRC | 12:33 | |
| *** jungleboyj has joined #openstack-tc | 12:33 | |
| *** gmann has joined #openstack-tc | 12:33 | |
| *** mnaser has joined #openstack-tc | 12:33 | |
| *** slaweq has joined #openstack-tc | 12:43 | |
| *** slaweq has quit IRC | 12:47 | |
| gmann | o/ | 13:31 |
| gmann | ttx: thanks, checking | 13:31 |
| gmann | ttx: +1, lgtm. thanks for updates and the chair idea. | 13:37 |
| *** jaosorior has quit IRC | 13:44 | |
| *** ricolin has joined #openstack-tc | 13:51 | |
| gmann | Board meeting started, if anyone would like to join- https://zoom.us/j/9574980705 | 14:02 |
| *** markvoelker has joined #openstack-tc | 14:16 | |
| *** iurygregory_ has joined #openstack-tc | 14:19 | |
| *** iurygregory has quit IRC | 14:19 | |
| *** iurygregory_ is now known as iurygregory | 14:19 | |
| mnaser | tc-members: it seems like openstack.org seems to list a few projects as "core functionality" -- https://www.openstack.org/software/ | 14:26 |
| mnaser | a few such as 'placement' and 'ironic' are listed, but i'm unsure how/why/where this is documented (im not against it, but i'm also curious) | 14:26 |
| mugsie | mnaser: if you ever find out, let me know, I have been asking for *years* | 14:27 |
| fungi | are you talking about the "landscape" map? | 14:28 |
| gmann | I will suggest to modify them from 'core' -> 'base services/stack' | 14:28 |
| mugsie | yeah | 14:28 |
| mnaser | fungi: correct, https://www.openstack.org/assets/openstack-map/openstack-map-v20190601.svg | 14:28 |
| mugsie | gmann: it isn't even that - Ironic is great, but not a base service for a lot of people | 14:28 |
| mugsie | same with swift | 14:28 |
| gmann | mugsie: yeah, with updating the services list also | 14:29 |
| mnaser | but also, nova could be not a base service for some people who want standalone ironic too | 14:29 |
| mnaser | so its tricky | 14:29 |
| gmann | or we just remove as interop now has adds-on program model | 14:29 |
| gmann | humm | 14:29 |
| fungi | https://opendev.org/osf/openstack-map/src/branch/master/openstack_components.yaml | 14:29 |
| fungi | so basically asking how/why certain projects get "map-bucket: openstack"? | 14:30 |
| mugsie | fungi: how is core defined in there? | 14:30 |
| fungi | oh, the bolding right? | 14:30 |
| mugsie | yeah | 14:30 |
| gmann | yeah, bold one | 14:30 |
| mnaser | right it says "Bold represents Core Functionality" | 14:30 |
| fungi | many of us keep telling osf marketing the bolding is a "bad idea" and just makes projects want to know why they're not worthy to be in bold | 14:30 |
| mugsie | (we have had this "core" fight for so long, I really wish we could remove it fully, and we baiscally had, until this map came out) | 14:31 |
| mugsie | and it is completely arbitrary - why is Ironic core, and not DNS or LBaaS? | 14:31 |
| fungi | i recommend proposing a change in gerrit to rip that out | 14:31 |
| mugsie | it isn't generated from that yaml though | 14:31 |
| fungi | provides a good hook to hang that argument on | 14:31 |
| fungi | oh, it isn't? | 14:31 |
| mugsie | nope | 14:32 |
| mnaser | yeah i dont see the script that generates it | 14:32 |
| fungi | ttx likely knows more | 14:32 |
| fungi | it looks to me like bold represents the oldest 8 openstack service projects and the placement service because it used to be part of nova | 14:34 |
| mugsie | ironic? | 14:34 |
| fungi | ironic was also part of nova, like placement | 14:35 |
| gmann | i also cannot find that in code. | 14:35 |
| fungi | but yeah, it's a specious definition of "core" | 14:35 |
| mugsie | might be worth a bug - the website bugs are on storyboard now, right? | 14:36 |
| fungi | osf/openstack-map repo might be, the website in general still takes bugs in lp | 14:36 |
| ttx | It's more of a marketing bolding to attract eyes to major projects | 14:37 |
| ttx | but it's true that we have not revisited it recently | 14:37 |
| mugsie | ttx: and we explictly wanted to get away from "core" due to its history in the project | 14:38 |
| ttx | Yeah, my draft map had "key" instead of "Core" | 14:38 |
| mugsie | but what is the definition for the bold? e.g. Ironic over Designate / Octavia / Manilla etc | 14:39 |
| * mugsie is getting deja vu | 14:39 | |
| fungi | any time you create a clear in-group, folks whose work is not in the in-group will want to know why | 14:39 |
| ttx | mugsie: it was mostly to match other marketing efforts. Like why is it nova CLI that is shows on openstack.org | 14:39 |
| fungi | (also it calls the cli an "api") | 14:40 |
| ttx | The map is a product of OSF marketing, it's a bit opinionated. But i agree that "core" was a poor choice (and advocated against it) | 14:40 |
| fungi | marketing is rarely "technically correct" it's a convenient fiction to help people think they understand something enough that they can move on | 14:40 |
| ttx | Putting all projects at the same level of size/bolding made the map difficult to navigate | 14:41 |
| mugsie | can we .. not do that on the home page for the project? | 14:41 |
| ttx | mugsie: not do what? | 14:41 |
| mugsie | use marketing reasoning for defining "core" or "major" funcitonality | 14:42 |
| fungi | mugsie: step #1 would be for the community to gain control over the openstack.org website. right now it's the osf marketing site which takes advantage of the openstack community to be able to have content to talk about | 14:42 |
| mugsie | I realise that Ironic is bolded due to the bare metal private cloud marketing push, but for a lot of our users, Octavia is a lot more "core" than Ironic | 14:43 |
| fungi | i personally would be much happier with a far less flashy and non-markety site with actually correct content, but there's a lot of history to overcome to get to that point | 14:43 |
| ttx | mugsie: it's tricky. You want the map to be navigable, and atract to most commonly-used projects (so see Nova more than Karbor) | 14:43 |
| mugsie | yeap - and I am pointing out projects that would be good to include for that definition | 14:43 |
| ttx | mugsie: I agree | 14:43 |
| ttx | We should had a few more, this has not been updated for years | 14:44 |
| ttx | (especially to avoid having another "what is core" discussion) | 14:44 |
| fungi | as for bug reports against osf/openstack-map, it's not got a corresponding project created on storyboard or launchpad, so the next best place to file a bug is probably https://bugs.launchpad.net/openstack-org/+filebug | 14:45 |
| ttx | but I agree that several projects have crossed that bar | 14:45 |
| ttx | So.. how about: | 14:45 |
| ttx | I add the bolding as a piece of information in the yaml | 14:45 |
| mugsie | (and this is with my TC hat fully off, and my "Graham has been pushing DNSaaS for too long hat) - DNS is a pretty core to a "DC Operating System" | 14:45 |
| ttx | then you can propose patches | 14:45 |
| mugsie | ttx: ++ | 14:45 |
| ttx | OK, I'll do that and let you know :) | 14:46 |
| ttx | FWIW some project teams do not appear, for the same reason -- the map is not the territory | 14:46 |
| gmann | +1. | 14:46 |
| fungi | also, how is the image generated? | 14:46 |
| ttx | so the goal is not to give everyone their corner of glory. But I agree that manila, or octavia, should probably be bold today | 14:47 |
| gmann | i was trying to find on openstackweb but could not, moving that to yaml will easy to update | 14:47 |
| fungi | hand-created taknig information from the yaml? | 14:47 |
| ttx | fungi: design team work | 14:47 |
| fungi | got it | 14:47 |
| ttx | I just check that the YAML matches :) | 14:47 |
| ttx | will do in osf/openstack-map | 14:47 |
| fungi | gmann: https://github.com/OpenStackweb/openstack-org/tree/master/software/images/map | 14:49 |
| fungi | the osf webdev team uses github (please don't ask why), so you could propose a pull request there i suppose | 14:49 |
| fungi | i haven't looked at the svg to see how directly editable it is though | 14:50 |
| *** slaweq has joined #openstack-tc | 14:52 | |
| gmann | fungi: thanks. svg is editable seems | 14:52 |
| ttx | seriously... openstack is hard to explain, and presenting everything at the same level just makes it super confusing. So yes it requires curation and marketing so that people can figure it out. Which is why the map is produced by OSF. The TC produces https://governance.openstack.org/tc/reference/projects/index.html | 14:53 |
| ttx | and yes, that means making choices | 14:54 |
| ttx | which the TC has explicitly said in the past it did not want to make | 14:54 |
| tbarron | if OSF marketing is conceiving of OpenStack chiefly as a "virtualization platform" (nova centric) rather than as open source Infrastructure as a Service | 14:54 |
| tbarron | that is a problem | 14:54 |
| ttx | tbarron: I don't think it does. Although I agree that the things in bold are due for a refresh | 14:55 |
| tbarron | that makes OpenStack more of an open source alternative to, say, VMWare rather than to AWS or GCP | 14:55 |
| tbarron | ttx: k | 14:55 |
| * tbarron fights that battle downstream | 14:56 | |
| ttx | the idea was to attract users to the most mature stuff (back then). Since the TC opted out of saying what is mature someone else had to | 14:56 |
| ttx | But I agree that several other components are at least as mature as the bold ones today | 14:56 |
| tbarron | with people who think that if workloads move off VMs to containers | 14:56 |
| tbarron | or VMs can themselves be run in containers | 14:57 |
| tbarron | there is no longer a need for OpenStack | 14:57 |
| ttx | anyway, I'll make sure the YAML exposes the "bolding" so that we can propose changes and discuss them in a review | 14:57 |
| tbarron | and that OpenStack itself can be replaced by some k8s based thing that doesn't have hard multi-tenancy baked into its API the way OpenStack does | 14:58 |
| * tbarron may be tangential to the main point her | 14:59 | |
| tbarron | here | 14:59 |
| tbarron | sorry | 14:59 |
| jungleboyj | o/ | 15:03 |
| * ricolin is paying attention on board meeting | 15:08 | |
| ttx | currently in the starlingx confirmation discussion | 15:10 |
| * jungleboyj is stuck in internal meetings. :-( | 15:10 | |
| * mugsie is the same as jungleboyj :( | 15:10 | |
| fungi | i'm quadruple-booked right now. this is a popular timeslot | 15:11 |
| ttx | jungleboyj: saw the Linux support announcement from Lenovo this morning, that's great news! | 15:11 |
| tbarron | yeah | 15:12 |
| jungleboyj | ttx: Isn't it? Though I would like to see support for more ThinkPad models, but hopefully that will follow. | 15:16 |
| *** dklyle has joined #openstack-tc | 15:17 | |
| ttx | it's already wider support than Dell's, if I understood the announce correctly. | 15:19 |
| *** slaweq has quit IRC | 15:20 | |
| smcginnis | Dell covers quite a bit. | 15:21 |
| jungleboyj | Oh it is on smcginnis ;-) | 15:23 |
| ttx | hmm... drink fight to solve it | 15:23 |
| smcginnis | :) | 15:23 |
| fungi | trying to install linux on a laptop usually is a good candidate for a drinking contest anyway | 15:24 |
| smcginnis | Any competition to see who can have better Linux support, the better all around. | 15:24 |
| ricolin | Starlingx just been approved as official | 15:25 |
| *** slaweq has joined #openstack-tc | 15:25 | |
| jungleboyj | :-) Dell was the first one to ship with Linux. | 15:25 |
| gmann | yeah as confirmed project | 15:25 |
| jungleboyj | Agreed. The more people shipping with Linux the better. | 15:26 |
| jungleboyj | I would love to see the X1 shipping with Linux. That would be my next laptop if they did it. | 15:26 |
| ttx | jungleboyj: you mean they don't?? I'll keep my XPS then | 15:28 |
| smcginnis | The new XPS 13 is very nice. ;) | 15:28 |
| smcginnis | And no more nose cam. | 15:28 |
| ttx | smcginnis: yes, I was tempted. But my current one is still working well enough | 15:29 |
| tbarron | Has there been recent Board discussion of making Membership involve more participation requirements? (topic currently is more about retention, which may be in tension with increased participation) | 15:29 |
| ttx | (and sees little use due to hmmm less travel) | 15:29 |
| *** slaweq has quit IRC | 15:30 | |
| gmann | tbarron: participation in contribution ? | 15:30 |
| smcginnis | tbarron: It has been brought up in multiple meetings, but I don't think anyone has come up with a good way to make it happen. | 15:30 |
| tbarron | gmann: yes | 15:30 |
| smcginnis | Especially when considering different ways to "contribute". | 15:30 |
| jungleboyj | They way I read the announcement was that it was just for the ThinkPad P series systems. | 15:30 |
| ttx | tbarron: we did discuss that in TC sessions at PTG | 15:30 |
| gmann | yeah, it would be good if it happen and i think it is much needed in current situation | 15:31 |
| tbarron | ttx: i'll read the record, was ovebooked and didn't lurk | 15:31 |
| tbarron | I push this issue with Daniel but it's all political | 15:31 |
| ttx | I personally would rather encourage users with significant deployments and business running on openstack to have an upstream person. That sounds more scalable | 15:31 |
| jungleboyj | ttx++ | 15:32 |
| ttx | we have much more users than OSF sponsors | 15:32 |
| ttx | currently some companies contribute by funding the OSF, some by giving infra resources, some by having developers upstream... Tying one to the others has proven difficult to enforce | 15:33 |
| ttx | and I'd rather have companies contribute because they want to, rather than because they need to | 15:34 |
| smcginnis | ++ | 15:34 |
| ttx | (in the latter case it's unlikely that their contribution would end up being very useful) | 15:34 |
| tbarron | ttx: all that makes sense but buying sponsorship without otherwise contributing bothers me a bit, maybe i should just lighten up | 15:35 |
| jungleboyj | ++ | 15:35 |
| jungleboyj | We had this discussion as part of the TC and it is difficult. | 15:35 |
| ttx | tbarron: oh I agree with you. But I'd say that with less hype, "buying sponsorship" has less benefits | 15:36 |
| jungleboyj | Providing financial input is important for the Foundation. It is weird that people would pay up and then not contribute. | 15:36 |
| jungleboyj | Lenovo wants to contribute, on the other hand, but money is hard to get. | 15:36 |
| jungleboyj | Corporations are very confusing. | 15:37 |
| tbarron | jungleboyj: oh we'll take laptops that run linux! | 15:37 |
| ttx | I agree that it would make more (business) sense that people who see a point in sponsoring money to OSF would also see business sense in contributing significantly upstream | 15:37 |
| jungleboyj | :-) I have tried to get you guys discounts there. | 15:37 |
| ttx | it's just hard to force it | 15:37 |
| tbarron | smcginnis: and a dell would be fine too | 15:37 |
| jungleboyj | I am happy to send my EPP info. | 15:37 |
| ttx | I'll take all and any free laptop. | 15:38 |
| jungleboyj | ttx: Ok, the EPP isn't that good. | 15:38 |
| ttx | jungleboyj: you need an influencer program! | 15:39 |
| fungi | the frogurt is also cursed | 15:39 |
| clarkb | ttx: it would be nice if we could influence the other direction and get double batteries, and removable battiers, and upgradeable sodimm slots, and no touchscreen on higher definitions monitors, etc back | 15:40 |
| fungi | also classing thinkpad keyboards | 15:42 |
| fungi | er, classic | 15:42 |
| zaneb | tbarron: OSF marketing has never not conceived of OpenStack chiefly as a nova-centric "virtualization platform" imho | 15:42 |
| zaneb | in large part I think because that's what the sponsors want | 15:42 |
| fungi | a common marketing tactic is to tell people what they already believe, because if you try to tell them something they believe is wrong, they're likely to just ignore you | 15:43 |
| ttx | zaneb: fwiw we are deploying a new homepage soon that insists on the "framework of components" | 15:43 |
| tbarron | zaneb: that is my fear. And that vision will IMO lead to obsolesence | 15:43 |
| jbryce | I'd disagree pretty strongly that we haven't conceived of it as more than virtualization | 15:44 |
| tbarron | everything must pass but I want a community dedicated to the 4 opens that develops open data center to survive | 15:44 |
| jbryce | When I'm not in the board meeting, I can point you to keynotes and materials going back to 2015 where we have talked about it as more than virtualization | 15:44 |
| fungi | the bolding ("core services") on the map we're putting front and center seems very virtualization-focused though | 15:45 |
| ttx | it's more mature-in-2017-focused really | 15:46 |
| ttx | or2015 | 15:46 |
| jbryce | Isn't only one of the bold things virtualization centric? | 15:46 |
| fungi | i think having ironic bold there helps | 15:47 |
| tbarron | jbryce: my broad brush is meant more as (hopefully) constructive provocation, as I see the virtualization-platform perception "out there" and hope the TC and Board will help correct | 15:47 |
| zaneb | 2015: https://twitter.com/zerobanana/status/575059167070785537 | 15:47 |
| tbarron | and I don't mean to be saying that efforts aren't there already | 15:48 |
| fungi | but the other bolded things there, if deployed without nova, aren't getting you much unless you add some of the non-bold things | 15:48 |
| *** e0ne has quit IRC | 15:48 | |
| *** e0ne has joined #openstack-tc | 15:49 | |
| tbarron | fungi: yeah, need compute instances and certainly there's been a lot of talk of late about running on baremetal | 15:49 |
| clarkb | fungi: swift keystone cinder and ironic all function without nova aiui and provide value to people doing that | 15:49 |
| jbryce | zaneb: who created that tag? | 15:49 |
| ttx | fungi: the bold stuffs was basically the trademark-things + ironic | 15:49 |
| tbarron | and that trademark heritage lives on | 15:50 |
| fungi | right, i think if we align the bolding with the interop trademark definition, that's far easier to explain | 15:50 |
| jbryce | As ttx said the bolding on the map could definitely use updates | 15:50 |
| ttx | tbarron: indeed. | 15:50 |
| fungi | at least then the answer to "how do i get my project to be important enough for bolding" is "add interop guidelines" | 15:50 |
| ttx | fungi: I'm not sure there is 100% overlap between what makes sense to overrepresent in the map, and what makes sense to trademark. | 15:51 |
| ttx | I'd rather take heat for being opinionated | 15:51 |
| gmann | I feel bold for interop guidelines projects might not be a complete things to bold. those can be with * or some other list. | 15:52 |
| fungi | ttx: luckily heat is already bolded ;) | 15:52 |
| ttx | I understand that the TC doesn't want to be in that role of picking what makes marketign sense | 15:52 |
| mugsie | and trademarked | 15:52 |
| gmann | true | 15:52 |
| ttx | that's the historic reasons why the map was created | 15:52 |
| ttx | because the TC would not make it | 15:52 |
| ttx | so yes, that means making choices | 15:52 |
| fungi | the tc at that time was not interested. perhaps the current tc is? it's not the same group of people | 15:53 |
| ttx | that the TC, as the representation of all openstack, is not well-equipped to make | 15:53 |
| *** markvoelker has quit IRC | 15:53 | |
| ttx | fungi: maybe. I think it's difficult to be the legal representation of everyone and have favorites | 15:53 |
| ttx | It also takes a marketign/design skillset that most on the TC don;t have | 15:54 |
| *** markvoelker has joined #openstack-tc | 15:54 | |
| ttx | (to make a map that conveys enough but not too much) | 15:55 |
| gmann | or just feed map from project.yaml (there if we do bold or core service anytime reflect the same in map). i mean no extra things than what TC defines in project.yaml | 15:55 |
| zaneb | jbryce: I believe technically nobody. it was renamed before being accepted by the TC, in part because of my objections. (but to your point, yes, the nova-centric view of openstack also enjoyed/enjoys widespread popularity in the technical community, and particularly the nova community) | 15:56 |
| ttx | every time we asked the TC to say what is "mature" or "important" they/we preefrred to not say | 15:56 |
| *** markvoelker has quit IRC | 15:56 | |
| fungi | it's probably fair to say that the tc has consistently objected to the idea that projects should be presented/compared/measured in that light, so it's not surprising they might object when it's done anyway | 15:57 |
| mugsie | constalations / starter-kits / sample configs were supposed to cover this | 15:57 |
| *** belmoreira has quit IRC | 15:57 | |
| mugsie | also what fungi said | 15:57 |
| *** rpittau is now known as rpittau|afk | 15:58 | |
| fungi | i can come to terms with the idea that the openstack tc and osf marketing have some hard disagreements on the best ways to present the public face of the project, and may simply have to agree to disagree | 15:59 |
| ttx | it's hard to explain "openstack" by starting with a list of 50 components equally presented | 16:00 |
| fungi | it makes sense that the sort of psychological tactics you need to apply to convince people of a particular point may not always be "technically correct" | 16:00 |
| ttx | fungi: and I think the TC has objected to doing it, not that someone else would do it. Same for "core" | 16:00 |
| mugsie | it *always* ends being used a gate - e.g. not integrated / core / tc-approved-release - no OS CLI / in tree tempest / dashboard / devstack / integrated gate for you | 16:01 |
| ttx | when we said that the approved release is basically whatever the board wants... we accepted that they would define it | 16:01 |
| fungi | a lot of the tc (and community at large) have expressed that the whole concept of "core" was terrible. but it does make something easier to explain (even if it's not correct conceptually for a lot of the community insiders) | 16:01 |
| tbarron | fungi: if the psychological tactics needed to present an open source alternative to aws or gcp require persuading people that to use it they need to be sysadmins on VMs | 16:03 |
| tbarron | that's when I have a problem | 16:03 |
| tbarron | i'm not anti-nova, quite the contrary | 16:03 |
| fungi | i do think that the "tc approved release" is an example of agreeing to disagree. the tc can say all of openstack is approved, even if the board insists on limiting it to a particular world view | 16:03 |
| tbarron | but to run a workload on open source IAAS one shouldn't have to know how to be a sysadmin | 16:04 |
| ttx | tbarron: I'm not sure where you get that impression... I mean yes Nova is in bold on the map, but so are Swift or Keystone | 16:06 |
| ttx | The main page shows Horizon and the openstack cli (yes, on a Nova example) | 16:06 |
| ttx | but does not mention VMs | 16:06 |
| clarkb | as a mostly happy VM user, I also think you need to be careful acting like my use case is a problem too | 16:07 |
| tbarron | clarkb: I don't think I am acting like your use case is a problem | 16:07 |
| ttx | if anything, the OSF has taken heat in the past for mentioning corner projects rather than presenting Nova all the time | 16:07 |
| clarkb | tbarron: the implication is that VMs are bad somehow and we should deemphasize that. Not necessarily that we should emphasize other use cases more | 16:08 |
| ttx | so... persuading people that to use it they need to be sysadmins on VMs ??? | 16:08 |
| clarkb | tbarron: I understand what you are trying to say, but it would be easy to interpret it the other way | 16:08 |
| tbarron | clarkb: I diddn't say VMs are bad or imply it | 16:08 |
| ttx | tbarron: not sure where your impression is raelly coming from | 16:09 |
| ttx | interested in pointers if you have some | 16:09 |
| clarkb | tbarron: I think the assertion that you need to be a sysadmin to use openstack while ignoring that other use cases are also highlighted puts a spotlight on that as if it were a problem | 16:09 |
| tbarron | ttx: yeah, let me regroup and try to articulate better cause it seems I am not communicating well. | 16:10 |
| clarkb | on that == using VMs | 16:10 |
| smcginnis | Definitely still a perception out in the wider world that OpenStack just means managing VMs, so if you don't want to use those old legacy things you should look for other projects. | 16:12 |
| clarkb | I think its fair to say a subset of use cases are highlighted and those don't reflect the entirety of openstack's usefulness. I think it is less fair to say the highlighting is nova (or VM) specific which imples to me that at least some of the problem is perceived to be nova (or VMs) | 16:12 |
| fungi | back to the original point, i do still feel like the mix of bold and normal font for project names on that diagram is distracting and makes it harder for me to interpret rather than easier. it's also entirely possible i don't consume information from these sorts of charts the way most people do. the callout mentioning that it represents "core functionality" also doesn't seem like it would convey much to | 16:13 |
| fungi | anyone coming at it fresh, unless they had a preconceived notion that some of what's listed there should be avoided | 16:13 |
| clarkb | I'm also a happy swift user completely outside of the context of nova (though at least one cloud we use also uses swift to manage VM image uploads) | 16:13 |
| tbarron | smcginnis says it better | 16:13 |
| smcginnis | hah | 16:13 |
| ttx | fungi: it was more useful back when we had a ton of very new projects | 16:13 |
| ttx | i.e. 2014-2015 | 16:14 |
| ttx | I agree that with most projects benig mature now it's less critical to direct visually people | 16:14 |
| gmann | what we need to do to change that perception ? just removing the bold things solve it? | 16:14 |
| smcginnis | tbarron: What's worse is a lot of people's assumption, including some OpenStack involved people, is that OpenStack is a free and incredibly more complicated alternative to vSphere. | 16:14 |
| fungi | yeah, maybe the bolding and "core" terminology has simply outlived whatever usefulness it once had | 16:15 |
| ttx | gmann: once I had the "bolding" data to the YAMl we can propose a chnage that removes all bolding and discuss that :) | 16:15 |
| ttx | s/had/add | 16:15 |
| *** iurygregory has quit IRC | 16:15 | |
| tbarron | smcginnis: agree. IMO having hard multi-tenancy built into the OpenStack APIs is a key distinguisihing feature. | 16:16 |
| gmann | yeah, i am worried if that still does not solve that perception of openstack being VM things only :) | 16:16 |
| tbarron | so when you use Swift on its own with S3 api even | 16:16 |
| clarkb | gmann: ya I'm beginning to think the concern is perceptions and we're attributing that entire problem to the project map | 16:16 |
| clarkb | I think the project map is more a symptom than a cause | 16:16 |
| tbarron | and I use it as a difft keystone identity | 16:16 |
| gmann | exactly | 16:16 |
| tbarron | there is common infra but we are ships in the night, isolated | 16:16 |
| tbarron | like on AWS S3, but open source, with 4 opens, community, etc. | 16:17 |
| tbarron | AND when we use nova vms, same kind of thing. | 16:17 |
| ttx | need to jump to another meeting | 16:18 |
| ttx | Will update the tc when I add things to openstack-map repo | 16:18 |
| gmann | ttx: thanks | 16:18 |
| tbarron | ttx: ty and i know i digressed :D | 16:18 |
| zaneb | smcginnis++ | 16:28 |
| zaneb | fwiw I don't consider the project map to be anything more than a minor symptom | 16:31 |
| * TheJulia ponders | 16:32 | |
| TheJulia | It is definitely a symptom, but we get that symptom via a lack of awareness and desire to comprehend as well as a desire to frame things into particular perspectives and ultimately be able to sell or offer that perspective as a solution. | 16:34 |
| *** markvoelker has joined #openstack-tc | 16:34 | |
| jungleboyj | smcginnis: ++ There is a significant perception issue. | 16:34 |
| jungleboyj | And many people see OpenStack as Infrastructure management only. The CNCF community also doesn't help avoid that situation. | 16:35 |
| fungi | i'm not sure it's their community originating that rhetoric, seems like it comes more from cncf's marketing arm within the lf | 16:39 |
| TheJulia | This feels much more like a issue of an incomplete feedback loop, and maybe it is just we didn't get feedback from the folks who tried to bend things to their will and threw up their hands and walked away before they gained the context they needed | 16:47 |
| *** ricolin has quit IRC | 16:48 | |
| zaneb | traditional cloud and CNCF represent competing approaches to the problem of how to deliver services (not compute) at wildly differing scales - i.e. from tiny users with 1 VM to Netflix or Snap | 16:48 |
| zaneb | traditional cloud - very large services (think SQS) with multitenancy | 16:48 |
| zaneb | CNCF - very small services, tightly packed (think NATS) | 16:48 |
| jungleboyj | Yes, very different approaches and problems. | 16:49 |
| zaneb | make no mistake, these are competitors | 16:49 |
| TheJulia | and different mind sets, and ultimately requirements that drive into those problems | 16:49 |
| zaneb | are they though? or just the standard centralised/decentralised tick-tock that we always see in computing? | 16:50 |
| TheJulia | Well, being told how the promised solving of all problems versus bending $thing to the requirements is one aspect I think everyone struggles with. Into that is the requirements that could be as simple (or painfully complex) as a contract line item. | 16:52 |
| TheJulia | I guess what I'm trying to say is it is not just computing and the software, but it is perceptions and use all the way back to the original reason why. There is always the desire to deliver solutions/solve problems. Variations in that begin to shift the way things are approached. Suddenly we reach "Hi, I need this software to fly a plane!" "Why?" "Because xyz person said so" "Why?" "Well... " | 16:54 |
| * jungleboyj doesn't want to imagine cloud software flying a plane. :-) | 16:56 | |
| jungleboyj | Though, with 5G, etc ... things like that may start to happen. | 16:56 |
| jungleboyj | Your car, powered by OpenStack. | 16:56 |
| TheJulia | powered by things that run on openstack | 16:57 |
| jungleboyj | :-) | 16:58 |
| fungi | skynet: powered by openstack | 16:58 |
| TheJulia | Sounds reasonable as long as it doesn't declare war on humans | 16:59 |
| jungleboyj | :-) | 17:00 |
| TheJulia | In fact, if it could work on building some colonies in the belt, that may be ideal.] | 17:02 |
| fungi | probably not until after luna and mars | 17:02 |
| jungleboyj | The plans to get to the moon again are impressive. | 17:03 |
| * jungleboyj also recommends watching 'Space Force' in this time where we all need a laugh. | 17:04 | |
| TheJulia | ++ | 17:04 |
| zaneb | anyway, we have largely failed to position OpenStack in the market as a (potential) traditional cloud in the sense I defined it above, and with the notable exception of Swift we haven't even really _built_ a serious competitor in that space | 17:12 |
| zaneb | now the tick-tock is headed in the decentralised direction anyway | 17:13 |
| zaneb | and kubernetes' extensible API has the great advantage that it can provide the orchestration for anything. people, correctly, hate having to mix multiple orchestration tools | 17:13 |
| zaneb | (if you think of k8s as a thing to manage containers you're missing the point. that's the #1 thing I have learned in the past year.) | 17:14 |
| zaneb | in conclusion, infrastructure management is all that is left, so it's no wonder people see us that way | 17:14 |
| zaneb | this should have been a blog post | 17:15 |
| smcginnis | Welcome to your TED talk? :) | 17:15 |
| zaneb | lol | 17:15 |
| zaneb | I also wrote a haiku about this but I'm not going to share it | 17:16 |
| fungi | i was hoping for a limerick | 17:17 |
| zaneb | (not joking about the haiku, for the record) | 17:17 |
| * jungleboyj is very curious about the haiku now. | 17:31 | |
| timburke | jungleboyj, fwiw, there are multiple companies working on self-driving cars and using swift to store their data :-) | 17:31 |
| jungleboyj | That is pretty cool. | 17:32 |
| clarkb | timburke: and zuul to test it | 17:33 |
| timburke | true! | 17:33 |
| fungi | some of those car companies even have employees who are zuul maintainers | 17:34 |
| fungi | and zuul-jobs core reviewers, and code contributors to the project | 17:35 |
| tbarron | timburke: i want self-driving swift and storage for my car | 17:53 |
| *** ralonsoh has quit IRC | 19:15 | |
| smcginnis | Some other folks here would probably be interested in this: https://opensource.org/StateOfTheSource | 19:24 |
| smcginnis | September through November is shaping up to be non-stop conferences. | 19:25 |
| *** slaweq has joined #openstack-tc | 19:49 | |
| *** slaweq has quit IRC | 20:26 | |
| *** slaweq has joined #openstack-tc | 20:31 | |
| *** slaweq has quit IRC | 20:35 | |
| *** markvoelker has quit IRC | 21:48 | |
| *** markvoelker has joined #openstack-tc | 21:49 | |
| *** markvoelker has quit IRC | 21:54 | |
| *** tkajinam has joined #openstack-tc | 22:46 | |
| *** markvoelker has joined #openstack-tc | 22:49 | |
| *** markvoelker has quit IRC | 22:54 | |
| *** tosky has quit IRC | 23:10 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
