| *** mordred has quit IRC | 00:49 | |
| *** tosky has quit IRC | 01:07 | |
| *** jamesmcarthur has quit IRC | 01:12 | |
| *** zaneb has quit IRC | 01:22 | |
| *** jamesmcarthur has joined #openstack-tc | 03:13 | |
| *** jamesmcarthur has quit IRC | 03:17 | |
| *** amrith has quit IRC | 03:32 | |
| *** whoami-rajat has joined #openstack-tc | 04:13 | |
| *** ricolin has joined #openstack-tc | 06:13 | |
| *** Luzi has joined #openstack-tc | 06:23 | |
| *** jamesmcarthur has joined #openstack-tc | 06:34 | |
| *** jamesmcarthur has quit IRC | 06:39 | |
| *** jaosorior has joined #openstack-tc | 07:41 | |
| *** eumel8 has joined #openstack-tc | 07:51 | |
| *** alexchadin has joined #openstack-tc | 07:51 | |
| *** alexchadin has quit IRC | 08:06 | |
| openstackgerrit | Alexander Chadin proposed openstack/governance master: Add assert:supports-upgrade to watcher https://review.openstack.org/619470 | 08:37 |
|---|---|---|
| *** tosky has joined #openstack-tc | 08:49 | |
| *** jpich has joined #openstack-tc | 09:01 | |
| openstackgerrit | Alexander Chadin proposed openstack/governance master: Add stable:follows-policy to watcher https://review.openstack.org/619481 | 09:04 |
| *** e0ne has joined #openstack-tc | 09:17 | |
| *** mrhillsman is now known as openlab | 09:19 | |
| *** openlab is now known as mrhillsman | 09:20 | |
| *** dtantsur|afk is now known as dtantsur | 09:54 | |
| *** cdent has joined #openstack-tc | 09:58 | |
| *** ianychoi has quit IRC | 10:25 | |
| *** ianychoi has joined #openstack-tc | 11:31 | |
| *** dims has quit IRC | 11:45 | |
| *** e0ne has quit IRC | 11:46 | |
| *** e0ne has joined #openstack-tc | 11:59 | |
| openstackgerrit | Jakub Sliva proposed openstack/governance master: Create repository for ceilometer-dashboard plugin https://review.openstack.org/619538 | 12:05 |
| *** cdent_ has joined #openstack-tc | 13:39 | |
| *** cdent has quit IRC | 13:39 | |
| *** cdent_ is now known as cdent | 13:39 | |
| *** dims has joined #openstack-tc | 14:14 | |
| *** whoami-rajat has quit IRC | 14:16 | |
| *** Luzi has quit IRC | 14:22 | |
| fungi | it's office hour, for those who are around anyway | 15:06 |
| cdent | oh hi | 15:07 |
| * fungi was exhausted and fighting jetlag/plague so passed out before the wednesday office hour, apologies | 15:07 | |
| mnaser | o/ | 15:08 |
| fungi | i liked the dburi envvar suggestion posted to the ml, btw. handy! | 15:08 |
| cdent | it's super handy | 15:10 |
| ttx | I'm here | 15:20 |
| evrardjp | that was very calm | 15:52 |
| evrardjp | fungi: are you still there? | 15:52 |
| fungi | yep | 15:53 |
| evrardjp | I was wondering, due to Dirk's review in https://review.openstack.org/#/c/611080/2, what really defines a 'supported' distro. I checked if there was anything in governance, there is vaguely something about it in the PTI part | 15:54 |
| evrardjp | I guess it must have been more with historical reasons, and ability to test in infra | 15:54 |
| evrardjp | but ... as Dirk mentioned it, it's possible to test openSUSE with devstack, so I am curious what defines the word 'popular' in here: https://governance.openstack.org/tc/reference/project-testing-interface.html#linux-distributions | 15:55 |
| evrardjp | fungi: any opinion, or historical reasons to share? Should I document those? | 15:56 |
| dhellmann | evrardjp : that list is maintained at https://governance.openstack.org/tc/reference/project-testing-interface.html#linux-distributions | 15:57 |
| fungi | it was based mostly on anecdotal reports and assumptions as to on which distros our users were installing openstack | 15:57 |
| evrardjp | fungi: that's what I was afraid of, tbh. | 15:58 |
| fungi | in years since validated by the periodic openstack user survey | 15:58 |
| scas | it's anecdotal, but from the chef camp, it was mostly based on what users wanted | 15:58 |
| fungi | but yes, possible that those early assumptions as to what users were running influenced later user choices | 15:58 |
| scas | not disparaging, but the only people who drove opensuse support in chef were employed by suse | 15:58 |
| evrardjp | well my employer would be super happy to be part of said list ;) | 15:58 |
| fungi | so it's hard to say for sure that the user survey validated those assumptions as it may have just as well been influenced by them | 15:58 |
| mugsie | yeah, people would turn up and help implement testing on $DISTRO, and I (in Designate land anyway) never had anyone try to add a SUSE gate | 15:59 |
| evrardjp | scas: that's true -- and I was told this wasn't continued (sadly) | 15:59 |
| scas | no, it was not. that's the nature of the bazaar, at the end of the day. | 15:59 |
| mugsie | are the opensuse images maintained in -infra? | 16:00 |
| evrardjp | mugsie: I am trying to change those mindsets to see if we should improve relationship with upstream projects. I mean I don't expect SUSE to be part of that list, if SUSE doesn't dedicate more resources to each of the projects to be 'popular' | 16:00 |
| evrardjp | mugsie: yes | 16:00 |
| scas | chef had a suse gate way, way long ago, in the hyped ages. when invested individuals moved away, bitrot set in and did what it naturally does | 16:01 |
| scas | there's no reason to think that one couldn't reintroduce it, with time and effort | 16:01 |
| scas | that's the kicker | 16:01 |
| evrardjp | I'd say that deployment projects isn't at the heart of the burning issue here -- I'm afraid not taking account of SUSE as 'supported distro' (without an explicit requirement of what 'supported distro is') would have a negative impact | 16:05 |
| evrardjp | I'd rather have a list of things to do for SUSE to become 'supported' :) | 16:05 |
| evrardjp | Nova is already listening to SUSE's input when bumping libvirt versions, I wonder why would it be different for python versions :D | 16:06 |
| evrardjp | call me a little naive at the same time :D | 16:06 |
| dhellmann | I think it's just a matter of no one pushing to have the list updated. I'm not even sure if we're clear about what the obligations & benefits of being on that list are, so we should start having the discussion about what it would take to add something new to the list and why someone might want to do that. | 16:08 |
| *** jamesmcarthur has joined #openstack-tc | 16:09 | |
| mugsie | I hope we would listen - py36 is the version that Leap would have supported, isn't it? | 16:09 |
| evrardjp | I will check for 15, as tumbleweed is probably different | 16:10 |
| mugsie | well, we would never have tumbleweed on the list | 16:10 |
| mugsie | :D | 16:10 |
| cmurphy | yes leap 15 is py36 | 16:10 |
| * dhellmann doesn't know what leap or tumbleweed are but likes whoever came up with the names | 16:10 | |
| mugsie | tumbleweed == rawhide I think? | 16:11 |
| mugsie | leap is the versioned release | 16:11 |
| cmurphy | tumbleweed is the rolling release, get it? ;) | 16:11 |
| evrardjp | cmurphy: this doesn't make sense at all! | 16:12 |
| fungi | it's worth pointing out that we do explicitly state what we mean by "supported" since it's perhaps a loaded term with a counter-intuitive definition for some | 16:13 |
| fungi | much of https://governance.openstack.org/tc/resolutions/20170620-volunteer-support.html applies to what we mean with the supported distros list | 16:14 |
| evrardjp | fungi: where is that explicit statement? Is that on the same page I linked? | 16:14 |
| evrardjp | okay | 16:15 |
| fungi | also, as a consumer, i'd be far more interested in "this distro supports openstack" than "openstack supports this distro" | 16:16 |
| mnaser | i think ownership of that fact is indeed a distro thing and not an openstack thing | 16:16 |
| evrardjp | I agree there | 16:16 |
| mnaser | i.e. we have support for opensuse in openstack-ansible but it's only cause folks showed up to do the work | 16:17 |
| mnaser | the hard part is keeping it consistently running | 16:17 |
| evrardjp | deployments projects are a tough case for this conversation :p | 16:17 |
| mnaser | right. but it's increased the workload on CI significantly, and causes a lot more test failures which end up being troubleshooted by another group of individuals | 16:18 |
| mnaser | which i think goes back to there is no set of arbitrary lists to be a "supported distro" | 16:19 |
| mnaser | show up, do the work and make it happen and then things will change when there is a clear visible set of users that deploy on said platform | 16:19 |
| evrardjp | there is an arbitrary list of supported distros though | 16:19 |
| scas | the most trodden paths tend to be laden with deployment projects. while them supporting distros was not an official signal, it did help shape the notion that There Are Two Distros | 16:20 |
| clarkb | it isnt arbitrary. the project grew out of a bunch of ubuntu de s (hence ubuntu support) and red hat has long been invol ed helping with testing too | 16:20 |
| evrardjp | mnaser: I am totally for that, there is no magic here | 16:20 |
| mnaser | yes, that list is because those distros have a significant amount of development that happens on them, developers from their parts and a few solid commercial offerings on top them | 16:20 |
| evrardjp | I really have to think multiple times before typing... that's a hard conversation for me. | 16:23 |
| evrardjp | maybe IRC isn't the right media for this conversation, at least for me. But at least historical reasons were explained, and that is in agreement with what I thought/expected. | 16:24 |
| cdent | I feel differently about this stuff. I think we should pick one distro, preferably one that doesn't sell OpenStack (so presumably Debian) and test on only it upstream and let the money oriented distros do third party CI | 16:24 |
| mnaser | cdent: what about deployment projeccts? :> | 16:24 |
| cdent | but as evrardjp said: [t 3ENT] | 16:25 |
| purplerbot | <evrardjp> call me a little naive at the same time :D [2018-11-22 16:06:30.723945] [n 3ENT] | 16:25 |
| scas | the typical deployment consists of One Distro. historically, that has been ubuntu, but there is money tied to that one | 16:25 |
| evrardjp | mnaser: what about it? | 16:25 |
| evrardjp | mnaser: deployment projects can choose | 16:25 |
| scas | debian would be the most likely of choices, regardless of my personal usage | 16:25 |
| scas | with portability being the fickle thing it is, it's hard to gauge 'support' if things are left to 'the community' | 16:26 |
| evrardjp | I really wanted to isolate this conversation of deployment projects, due to the nature of deployment projects (which could be opinionated), where the non-deployment projects should not be opinionated on their platform (but be more agnostic of the OS) | 16:26 |
| fungi | yeah, i (perhaps selfishly) would love if we could just standardize on testing on debian | 16:26 |
| mnaser | fungi: me too, honestly | 16:27 |
| fungi | problem is we have a lot of functionality in openstack which calls into other parts of the operating system, and some of those differ across distros, so knowing we're not breaking those things is helpful | 16:27 |
| clarkb | my only concern with debian is stable is too stable so we'd have to find the unstable/testing balance | 16:27 |
| evrardjp | I don't know for my employer, but that sounds a more 'neutral' stance for me. | 16:27 |
| evrardjp | clarkb: is it still the case now? | 16:28 |
| *** e0ne has quit IRC | 16:28 | |
| *** lbragstad has joined #openstack-tc | 16:28 | |
| scas | aren't the debian packages built against sid? | 16:28 |
| mugsie | fungi: yeah, differiong versions of dependancies can be a problem | 16:28 |
| evrardjp | it's also not really bad to be stable :) | 16:28 |
| mugsie | evrardjp: well, for things like python testing it can be | 16:29 |
| clarkb | evrardjp: it is if your goal is to test the future of things | 16:29 |
| fungi | clarkb: debian/unstable is in fact quite stable as well (at least as stable as ubuntu for me). it's a rolling distro with some safeguards and an increasing amount of automated testing (though not to the degree of suse's rolling release thing i'm sure) | 16:29 |
| ttx | fungi, clarkb, dhellmann: shouldn't you be turkey-eating right now ? | 16:29 |
| clarkb | ttx yes I sat down for tea | 16:29 |
| cmurphy | it's too early for turkey | 16:29 |
| ttx | cmurphy: heresy | 16:30 |
| fungi | ttx: i will at some point, but it's not a holiday i celebrate particularly | 16:30 |
| evrardjp | sorry I pulled you all for that conversation where you could be with family ALL DAY | 16:30 |
| mnaser | no turkey in canada | 16:30 |
| mnaser | just really cold. | 16:30 |
| mugsie | or supporting (in my case) never versions of powerdns or novas, libvirt | 16:30 |
| ttx | mnaser: only poutine | 16:30 |
| evrardjp | mnaser: I didn't know frozen turkey was a thing | 16:30 |
| scas | 08:30 is too early for anything but tea or coffee | 16:30 |
| mnaser | mmm that sounds good, i haven't had that in a while | 16:30 |
| * mugsie now wants poutine | 16:30 | |
| * evrardjp too | 16:30 | |
| evrardjp | mnaser: can you really call yourself canadian then? | 16:31 |
| mnaser | ill post photos for y'all | 16:31 |
| dhellmann | ttx: lunch is still cooking, but yeah, I'm failing at holidaying | 16:31 |
| evrardjp | I don't know what the requirements are for 'supported canadian' | 16:31 |
| evrardjp | :D | 16:31 |
| *** ricolin has quit IRC | 16:31 | |
| clarkb | fungi: that is good to know. | 16:32 |
| evrardjp | we can still talk about distros and testing after turkey, don't worry dhellmann | 16:32 |
| fungi | clarkb: i run unstable on all my desktops/portables and latest stable on all my servers. testing both would probably not be very hard | 16:32 |
| evrardjp | clarkb: stable for stable/ branches sounded attractive to me, while unstable on master | 16:33 |
| *** simon-AS559 has joined #openstack-tc | 16:33 | |
| evrardjp | but I am feeding the troll there :p | 16:33 |
| fungi | the most unstable thing about debian/unstable is that it's allowed to have transitional package conflicts, so there may be points in time where you can't successfully install some set of packages. that's where using testing instead of unstable might pay off for ci | 16:33 |
| fungi | testing is mostly the result of holding packages from unstable out until they can be coinstalled and have their dependencies met | 16:34 |
| fungi | except that it undergoes freezes leading up to stable releases, making it a little inconsistently-paced for a few months every couple years | 16:34 |
| clarkb | as a happy ubuntu and debian and suse user I'm probably the erong person to ask too. I dont get as religious about this and use $tool that gets the job done. | 16:36 |
| clarkb | debian staboe is great for an irc bouncer :) | 16:36 |
| fungi | also, the pace of debian stable releases has been roughly the same as that of ubuntu lts releases over the past ~decade | 16:37 |
| clarkb | fungi true but the package lag is longer iirc. Ubuntu is less conservative pre release | 16:37 |
| fungi | if you look at the debian release history it's around 2-2.5 years between each, just not scheduled | 16:37 |
| clarkb | so ehile cadence is similar the ubuntu packages tend to be newer | 16:37 |
| scas | two decades ago, i might have become incensed at choices of distro. i have a healthy mixture in my household, sans suse, but i ran tumbleweed on a workstation for a time | 16:38 |
| fungi | right, ubuntu grabs a bunch of packages out of debian testing/unstable, tests them together, mashes some of their own packages/patches in, and calls it a release as of some date | 16:38 |
| evrardjp | OMG I opened pandora's box. | 16:39 |
| evrardjp | Please don't close it on hope. | 16:39 |
| *** lbragstad has quit IRC | 16:39 | |
| fungi | debian's release process does attempt to boil the ocean a bit with requiring release-critical bug counts below a particular threshold before the release can be called | 16:39 |
| mugsie | evrardjp: the big debates - editors, distros, tabs vs spaces, CLI vs GUI and lanugages :P | 16:40 |
| * mugsie is in a #development channel with the topic set to "Monday: Spaces vs. Tabs, Tuesday: CLI vs GUI, Wednesday: IDEs, Thursday: Language Wars, Friday: Coffee, Saturday: Developer vs. Engineer" | 16:41 | |
| tosky | at least one of them it's settled in this community (spaces!) | 16:41 |
| * fungi notes that channel uses a 6-day week | 16:41 | |
| scas | i think it's healthy to trot some of them out from time to time. especially as many involved have had a significant history that involve different perspectives | 16:41 |
| mugsie | fungi: the theory is that sunday is a day that none of us should be on chat :) | 16:42 |
| * fungi misses 3-space tabs | 16:42 | |
| evrardjp | mugsie: I didn't want to start this -- it was more about *hey, what can we do so you don't forget us!* -- but ofc it's vim. | 16:42 |
| mugsie | but that is just a theory | 16:42 |
| tosky | mugsie: almost everyone - let's say "one of the 7 days" | 16:43 |
| evrardjp | mugsie: that channel sounds interesting, probably FOSDEM presentation material | 16:44 |
| mugsie | eh ... more like a load of bitter Irish people fighting in slack :P | 16:44 |
| evrardjp | (friday only) | 16:44 |
| evrardjp | hahaha | 16:44 |
| clarkb | from the infra teams perspective we strive to have images for what ope stack has asserted it needs and the other images are largely brought in via community help (suse gentoo debian). If openstack says it wants debian we can transition to that Likely will require involvement from people that understand things like the half systemd/sysv init transition state | 16:46 |
| fungi | it's really not in a half-transition state | 16:47 |
| clarkb | it is on my debian machine :/ | 16:48 |
| fungi | many packages do still have only sysvinit scripts and no systemd unit files, but by default they use systemd's sysv compat functionality to call those with no problem | 16:48 |
| clarkb | ya that causes problems if you want to modify unit behavior | 16:48 |
| clarkb | its fine if you dont | 16:48 |
| fungi | got it. yes you end up needing to edit initscripts in some cases rather than unit files | 16:49 |
| fungi | it's been a lazy transition, letting package maintainers add systemd unit files whenever they get around to writing them (or when someone contributes a patch to add them) | 16:50 |
| *** simon-AS5591 has joined #openstack-tc | 16:51 | |
| *** simon-AS559 has quit IRC | 16:51 | |
| scas | without money-driven development, some individuals can find it impossible to justify exchanging time and effort otherwise, especially in projects that still need it for long-term continuity. i've been witnessing one such case unfold upstream from where i sit | 16:52 |
| clarkb | Oh I dont think its a bug that debian does it as much as something ti address in any porting | 16:54 |
| fungi | yup | 16:54 |
| clarkb | iirc we drop unit overrid fioes for ra dom services an if those are sysv init scripts on debian we'll ave to soet that out | 16:54 |
| clarkb | wow typing hard | 16:54 |
| mnaser | clarkb: on a super side note | 16:55 |
| evrardjp | I still understood you | 16:55 |
| mnaser | can we have rhel 8 in infra or is there $reasons behind why we cant? | 16:55 |
| evrardjp | but please change your keyboard | 16:55 |
| evrardjp | :p | 16:55 |
| evrardjp | mnaser: rhel? | 16:56 |
| clarkb | mnaser: the last time we tried to have rhel I think we all decided we'd never do that again | 16:56 |
| clarkb | mnaser it was impossiboe to figure out the billing/lice sing between us,clouds, and red hat | 16:56 |
| clarkb | so we used centos instead | 16:56 |
| mnaser | right, rhel 8 *only* until centos 8 is out | 16:56 |
| mnaser | just so deployment projects can start having some outlook on how things might break | 16:57 |
| clarkb | that would still require licensing right? | 16:57 |
| clarkb | that is the hang up | 16:57 |
| scas | licensing still applies to the beta | 16:57 |
| mnaser | i mean afaik you can install but not update with rhel? i dunno, never dealt with it | 16:57 |
| scas | free* licensing | 16:57 |
| evrardjp | mnaser: couldn't you try in your ci systems? :p | 16:57 |
| *** diablo_rojo has joined #openstack-tc | 16:58 | |
| mnaser | we could, but i try to avoid doing things in a closed box | 16:58 |
| mnaser | i can imagine other folks benefitting from this too | 16:58 |
| mnaser | https://developers.redhat.com/rhel8/getrhel8/ "no cost" | 16:58 |
| evrardjp | well that's until centos8 is out ofc :) | 16:58 |
| scas | it's licensed under a developer account, which are single entitlement installs | 16:59 |
| mugsie | no cost - but do you have to register each boot of an instance with a key? | 16:59 |
| mnaser | mugsie: that's a good question that outside holiday we surely have enough redhat people to find answers for :) | 16:59 |
| mugsie | yeah, that doies seem like a monday question | 17:00 |
| scas | doesn't look to be key-driven, but i haven't personally booted a vm to see | 17:00 |
| scas | i would presume not, based on history | 17:00 |
| clarkb | "download the iso fioe" ugh | 17:01 |
| fungi | to clarify, the last time we tried it the cloud providers said we couldn't boot rhel in their instances without arranging licensing through the cloud service provider because they were under contract with rh to track licenses for them, and we couldn't manage to get rh to comp the correct licenses to the providers to supply to our instances | 17:01 |
| clarkb | why are isos still a thing I dont even have a spinning disk drive anymore | 17:01 |
| fungi | these days "iso" is how you spell "usb image with a legacy bootloader" | 17:02 |
| clarkb | fungi also I think the cloud provider has to run rhel too or something | 17:02 |
| clarkb | so hp couldnt rheo at all | 17:02 |
| mugsie | clarkb: for us foolish mortals who have to boot via ISO uploaded to iDRACs :'( | 17:02 |
| fungi | clarkb: oh, yes at least at the time rh wouldn't provide licenses to rhel instances running on !rhel hosts. i can't remember if they've relaxed that in more recent years | 17:03 |
| mugsie | bacause "thats how we have always done it" | 17:03 |
| evrardjp | OK I have to go everyone. Enjoy your holidays for those who have some :) | 17:03 |
| clarkb | mugsie: I once had a terrible install method for freebsd + zfs root that included nfs mounting the install media remotely. Idrac + iso is maybe not the worst thing | 17:04 |
| clarkb | it is just odd to me that everyone has an image builder noe but the official install media for rhel and xenserver etc is still an iso | 17:05 |
| tosky | but there are also official cloud images | 17:08 |
| mnaser | for rhel? | 17:09 |
| tosky | afaik yes, for everyone | 17:12 |
| fungi | https://access.redhat.com/solutions/894123 | 17:12 |
| fungi | though i'm not about to register an account with red hat to find out the answer to that | 17:12 |
| tosky | I can tell that there are for sure for all released versions, alongside ISO images | 17:13 |
| tosky | I'm just not sure about RHEL8 beta, but I'd be surprised if they are not available as well | 17:13 |
| fungi | the rhel downloads pages i found for them also seem to require an account/registration so can't be obtained anonymously | 17:14 |
| fungi | clarkb: my searching around suggests that if you manage to retrieve the rhel .raw diskimages the kickstart scripts they used to build those are included in them | 17:16 |
| tosky | same rules as for ISOs | 17:16 |
| fungi | tosky: yep, but i already get spammed/tracked by enough people trying to sell me things i don't want | 17:16 |
| fungi | so i'm not about to hand my info over to rh just to find out | 17:17 |
| mugsie | no rhel 8 qcow download from what I can see | 17:18 |
| mugsie | I can get them for 7.x, but 8 doesn't seem to be in my list | 17:18 |
| *** jpich has quit IRC | 17:20 | |
| tosky | ok, I'm surprised :/ | 17:23 |
| *** e0ne has joined #openstack-tc | 17:25 | |
| mugsie | I can't dig too deep thoough, the RH site is having auth server issues it seems | 17:26 |
| cdent | I find it truly bizarre that we are even talking about trying to use something upstream that requires navigating auth and licenses | 17:27 |
| cdent | how does that even map? | 17:27 |
| mugsie | I am not :) | 17:27 |
| mugsie | I just got interested in finding qcows | 17:28 |
| cdent | moo | 17:28 |
| cdent | sorry | 17:28 |
| cdent | qmoo | 17:28 |
| * mugsie lol'd | 17:29 | |
| mugsie | even if we could get rhel working in our clouds, we need systems that others can use to replicate failures, and I am not a fan of asking a dev to get a rhel licence to debug why a devstack gate is failing | 17:31 |
| *** dtantsur is now known as dtantsur|afk | 17:35 | |
| *** jamesmcarthur has quit IRC | 17:35 | |
| fungi | yes, 100% with you on all of the above | 17:36 |
| fungi | this is why we use centos as a stand-in | 17:36 |
| fungi | we're told that a significant number of "enterprise" openstack deployments are on rhel, and centos is a "close enough" proxy for testing that we don't break on rhel | 17:38 |
| scas | to get back to the crux of the matter, i think the verbiage on supportability can be softened / expanded to not call out distro names specifically | 17:38 |
| scas | or call out a list ending with 'etc.', something that doesn't say There Are Two Distros | 17:39 |
| *** jamesmcarthur has joined #openstack-tc | 17:39 | |
| fungi | more distro-agnostic wording would be nice if only to reduce the number of arguments based solely on people/companies who are left out of the list vying for inclusion. though we still need some guidance as to what projects should be testing | 17:39 |
| scas | regardless of the opinions of deployment projects, one could use that as a definitive resource for justifying usage | 17:39 |
| scas | you can't please everyone all the time. someone will always be left out | 17:40 |
| fungi | the problem with publishing *any* sort of list is that there are entities who, when not included, will want to know what's involved in getting added | 17:40 |
| fungi | this is, for example, why the openstack vmt doesn't publish its downstream stakeholders list. we're fine with anyone who has a legitimate need requesting addition, but we don't want requests to be added just so they're not missing from some published list of names | 17:41 |
| fungi | as soon as you publish any list of competitors, it becomes marketing material (whether you want it to be or not) | 17:42 |
| scas | sure, that's a risk. in my time as lorax, i've had individuals approach me about including support for a given component. it's a balance over what the long-term caretakers can handle versus what the short-term opportunists want for some blog spam | 17:42 |
| *** e0ne has quit IRC | 17:49 | |
| *** jamesmcarthur has quit IRC | 17:56 | |
| *** jamesmcarthur has joined #openstack-tc | 18:00 | |
| *** e0ne has joined #openstack-tc | 18:07 | |
| *** jamesmcarthur has quit IRC | 18:08 | |
| *** e0ne has quit IRC | 18:15 | |
| *** jamesmcarthur has joined #openstack-tc | 18:45 | |
| *** simon-AS5591 has quit IRC | 18:49 | |
| *** jamesmcarthur has quit IRC | 18:53 | |
| *** jamesmcarthur has joined #openstack-tc | 18:57 | |
| *** jamesmcarthur has quit IRC | 19:02 | |
| *** diablo_rojo has quit IRC | 19:59 | |
| *** simon-AS559 has joined #openstack-tc | 20:50 | |
| *** simon-AS559 has quit IRC | 20:55 | |
| *** simon-AS559 has joined #openstack-tc | 20:55 | |
| *** lbragstad has joined #openstack-tc | 21:54 | |
| *** lbragstad has quit IRC | 23:23 | |
| *** tosky has quit IRC | 23:31 | |
| *** jamesmcarthur has joined #openstack-tc | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!