| *** walharthi has quit IRC | 00:16 | |
| *** walharthi has joined #openstack-glance | 00:33 | |
| *** walharthi has quit IRC | 00:45 | |
| *** catintheroof has quit IRC | 00:47 | |
| *** catintheroof has joined #openstack-glance | 00:48 | |
| *** catintheroof has quit IRC | 00:52 | |
| *** rrecio_ has quit IRC | 00:53 | |
| *** haplo37 has quit IRC | 01:20 | |
| *** haplo37 has joined #openstack-glance | 01:25 | |
| *** tsymanczyk has quit IRC | 01:34 | |
| *** lujinluo has joined #openstack-glance | 01:49 | |
| *** markvoelker has quit IRC | 01:50 | |
| *** kfox1111 is now known as kfox1111_away | 01:51 | |
| *** markvoelker has joined #openstack-glance | 02:19 | |
| *** mnaser has quit IRC | 02:24 | |
| openstackgerrit | melissaml proposed openstack/glance-specs: Fix a typo in artifact-repository.rst https://review.openstack.org/397482 | 02:26 |
|---|---|---|
| *** walharthi has joined #openstack-glance | 02:35 | |
| *** namnh has joined #openstack-glance | 02:37 | |
| *** mnaser has joined #openstack-glance | 02:37 | |
| *** namnh has quit IRC | 02:40 | |
| *** namnh has joined #openstack-glance | 02:40 | |
| *** links has joined #openstack-glance | 02:49 | |
| *** bkopilov_ has quit IRC | 02:57 | |
| *** namnh has quit IRC | 03:04 | |
| *** jamielennox is now known as jamielennox|away | 03:07 | |
| *** GB21 has joined #openstack-glance | 03:08 | |
| *** GB21 has quit IRC | 03:19 | |
| *** jamielennox|away is now known as jamielennox | 03:21 | |
| *** TravT has joined #openstack-glance | 03:25 | |
| *** TravT has quit IRC | 03:27 | |
| *** TravT_ has joined #openstack-glance | 03:27 | |
| *** TravT_ has quit IRC | 03:27 | |
| *** TravT has joined #openstack-glance | 03:28 | |
| *** TravT has quit IRC | 03:33 | |
| *** udesale has joined #openstack-glance | 03:42 | |
| *** nicolasbock has quit IRC | 03:54 | |
| *** TravT has joined #openstack-glance | 03:57 | |
| *** TravT has quit IRC | 04:08 | |
| *** TravT has joined #openstack-glance | 04:08 | |
| *** TravT has quit IRC | 04:09 | |
| *** TravT has joined #openstack-glance | 04:09 | |
| *** jamielennox is now known as jamielennox|away | 04:15 | |
| *** TravT has quit IRC | 04:15 | |
| *** jamielennox|away is now known as jamielennox | 04:17 | |
| *** bkopilov_ has joined #openstack-glance | 04:17 | |
| *** groen692 has joined #openstack-glance | 04:33 | |
| *** prateek has joined #openstack-glance | 04:38 | |
| *** jamielennox is now known as jamielennox|away | 04:40 | |
| *** ratailor has joined #openstack-glance | 05:00 | |
| *** jamielennox|away is now known as jamielennox | 05:03 | |
| *** groen692 has quit IRC | 05:08 | |
| *** walharthi has quit IRC | 05:17 | |
| *** pdeore has joined #openstack-glance | 05:23 | |
| *** bhagyashris has quit IRC | 05:56 | |
| *** lujinluo has quit IRC | 06:15 | |
| *** lujinluo has joined #openstack-glance | 07:12 | |
| *** bhagyashris has joined #openstack-glance | 07:17 | |
| *** tshefi has joined #openstack-glance | 07:22 | |
| *** rcernin has quit IRC | 07:23 | |
| *** belmoreira has joined #openstack-glance | 07:32 | |
| *** rvasilets___ has joined #openstack-glance | 07:36 | |
| *** exploreshaifali has joined #openstack-glance | 07:38 | |
| openstackgerrit | Jaspreet Singh Rawel proposed openstack/python-glanceclient: Null value of image property name should not allowed https://review.openstack.org/396084 | 07:54 |
| openstackgerrit | Jaspreet Singh Rawel proposed openstack/python-glanceclient: Show disk format vhdx in help https://review.openstack.org/395559 | 08:08 |
| *** pcaruana has joined #openstack-glance | 08:22 | |
| *** e0ne has joined #openstack-glance | 08:24 | |
| *** e0ne has quit IRC | 08:25 | |
| *** zul has quit IRC | 08:33 | |
| *** zul has joined #openstack-glance | 08:44 | |
| *** markvoelker has quit IRC | 08:49 | |
| *** lujinluo has quit IRC | 08:59 | |
| *** zzzeek has quit IRC | 09:00 | |
| *** zzzeek has joined #openstack-glance | 09:00 | |
| *** lin_yang has quit IRC | 09:13 | |
| *** mvk has quit IRC | 09:13 | |
| *** mvk has joined #openstack-glance | 09:15 | |
| *** openstackgerrit has quit IRC | 09:18 | |
| *** openstackgerrit has joined #openstack-glance | 09:18 | |
| *** clsacramento has joined #openstack-glance | 09:21 | |
| *** rvasilets___ has quit IRC | 09:38 | |
| *** udesale has quit IRC | 09:45 | |
| *** markvoelker has joined #openstack-glance | 09:49 | |
| *** markvoelker has quit IRC | 09:55 | |
| *** udesale has joined #openstack-glance | 09:56 | |
| *** exploreshaifali has quit IRC | 10:10 | |
| *** e0ne has joined #openstack-glance | 10:19 | |
| *** udesale has quit IRC | 10:53 | |
| *** foutatoro has joined #openstack-glance | 11:04 | |
| foutatoro | Hi all, I would like to know how to disable cloud-init config at the launch of instances in order to get them ready as fast as possible? | 11:13 |
| *** mvk has quit IRC | 11:16 | |
| *** houming has joined #openstack-glance | 11:40 | |
| *** pdeore has quit IRC | 11:42 | |
| *** nicolasbock has joined #openstack-glance | 11:42 | |
| *** links has quit IRC | 11:43 | |
| *** foutatoro has quit IRC | 11:45 | |
| *** smatzek has joined #openstack-glance | 11:47 | |
| *** mvk has joined #openstack-glance | 11:47 | |
| *** smatzek has quit IRC | 11:49 | |
| *** houming has quit IRC | 11:49 | |
| *** markvoelker has joined #openstack-glance | 11:51 | |
| *** markvoelker has quit IRC | 11:55 | |
| *** abhishek_k has joined #openstack-glance | 12:02 | |
| *** catintheroof has joined #openstack-glance | 12:28 | |
| *** bkopilov_ has quit IRC | 12:31 | |
| *** walharthi has joined #openstack-glance | 12:50 | |
| *** walharthi has quit IRC | 12:54 | |
| *** e0ne has quit IRC | 12:56 | |
| *** e0ne has joined #openstack-glance | 12:56 | |
| *** pdeore has joined #openstack-glance | 12:59 | |
| *** ducttape_ has joined #openstack-glance | 13:08 | |
| *** ratailor has quit IRC | 13:12 | |
| *** pdeore has quit IRC | 13:12 | |
| *** edmondsw has joined #openstack-glance | 13:18 | |
| *** walharthi has joined #openstack-glance | 13:22 | |
| *** croelandt has joined #openstack-glance | 13:28 | |
| *** markvoelker has joined #openstack-glance | 13:28 | |
| *** prateek has quit IRC | 13:34 | |
| *** ducttape_ has quit IRC | 13:35 | |
| *** ducttape_ has joined #openstack-glance | 13:35 | |
| *** ducttape_ has quit IRC | 13:40 | |
| openstackgerrit | melissaml proposed openstack/glance-specs: Fix a typo in artifact-repository.rst https://review.openstack.org/397482 | 13:55 |
| *** prateek has joined #openstack-glance | 14:00 | |
| *** udesale has joined #openstack-glance | 14:00 | |
| croelandt | jokke_: don't we have a meeting now? | 14:05 |
| rosmaita | croelandt: i think i was going to re-start next week after O-1 | 14:07 |
| rosmaita | but, i'm here, so we can meet anyway | 14:08 |
| croelandt | oh right | 14:08 |
| croelandt | ok, I only read half of my mails | 14:08 |
| croelandt | good, I have a ton of stuff on the way right now :D | 14:08 |
| rosmaita | that's ok, i;m way behind | 14:08 |
| rosmaita | ok, let's meet next week, is 14:00 UTC still good for you? | 14:09 |
| openstackgerrit | Merged openstack/glance-specs: Fix a typo in artifact-repository.rst https://review.openstack.org/397482 | 14:09 |
| *** jaosorior has joined #openstack-glance | 14:11 | |
| *** ducttape_ has joined #openstack-glance | 14:13 | |
| *** pcaruana has quit IRC | 14:13 | |
| *** ducttape_ has quit IRC | 14:18 | |
| croelandt | rosmaita: yep, ok with me | 14:29 |
| *** bkopilov_ has joined #openstack-glance | 14:29 | |
| rosmaita | croelandt: ok, cool | 14:32 |
| *** jaosorior has quit IRC | 14:36 | |
| *** ducttape_ has joined #openstack-glance | 14:40 | |
| *** ducttape_ has quit IRC | 14:52 | |
| *** groen692 has joined #openstack-glance | 14:54 | |
| *** udesale has quit IRC | 14:54 | |
| *** walharth_ has joined #openstack-glance | 14:57 | |
| *** walharthi has quit IRC | 14:58 | |
| *** exploreshaifali has joined #openstack-glance | 14:58 | |
| *** ducttape_ has joined #openstack-glance | 15:01 | |
| *** jdillaman1 has quit IRC | 15:05 | |
| *** groen693 has joined #openstack-glance | 15:05 | |
| *** groen692 has quit IRC | 15:06 | |
| *** groen693 is now known as groen692 | 15:06 | |
| *** prateek has quit IRC | 15:07 | |
| *** exploreshaifali has quit IRC | 15:10 | |
| *** jdillaman has joined #openstack-glance | 15:10 | |
| sigmavirus | rosmaita: I've got our ducks in a row: https://review.openstack.org/397793 | 15:24 |
| openstackgerrit | Tomislav Sukser proposed openstack/glance: IPv6 fix in Glance for malformed URLs. https://review.openstack.org/397794 | 15:24 |
| openstackgerrit | Tomislav Sukser proposed openstack/glance: IPv6 fix in Glance for malformed URLs. https://review.openstack.org/397797 | 15:29 |
| openstackgerrit | Hemanth Makkapati proposed openstack/glance: Port Glance Migrations to Alembic https://review.openstack.org/382958 | 15:30 |
| rosmaita | sigmavirus: ty | 15:33 |
| rosmaita | jokke_: when you have a minute, can you let me know what you think of my reply to you on https://review.openstack.org/#/c/396919/1 ? Thanks! | 15:48 |
| *** openstackgerrit has quit IRC | 15:48 | |
| *** openstackgerrit has joined #openstack-glance | 15:48 | |
| *** tsymanczyk has joined #openstack-glance | 15:49 | |
| tsymanczyk | jokke_ : could you please clarify your last comment on this patchset? https://review.openstack.org/#/c/369110/ "Now when the translation files are not touched anymore that is my only major concern about this change."? I fixed that yesterday, as far as I know. What is it that you still don't like? The .po aren't touched anymore. | 15:51 |
| sigmavirus | tsymanczyk: jokke_'s concerns are on rosmaita's spec update (dealing with "shared" being the default | 15:51 |
| tsymanczyk | sigmavirus : that part of his comment, sure. i'm just confused about the quoted part. | 15:52 |
| openstackgerrit | Hemanth Makkapati proposed openstack/glance: Testing changes required with Alembic move https://review.openstack.org/397409 | 15:52 |
| tsymanczyk | it reads to me as if i did not fix the translation files correctly | 15:52 |
| rosmaita | tsymanczyk: i think he means he's fine with what you have now | 15:52 |
| rosmaita | except that he disagrees about the 'shared' as default | 15:52 |
| tsymanczyk | ok. just making sure there's nothing waiting on me. | 15:53 |
| *** rrecio_ has joined #openstack-glance | 15:53 | |
| sigmavirus | I interpret that the same as rosmaita | 15:53 |
| rosmaita | tsymanczyk: thanks for checking | 15:53 |
| rosmaita | sigmavirus: thanks for the confirmation! | 15:54 |
| sigmavirus | tsymanczyk: I'll stop the world and wait on you | 15:54 |
| rosmaita | isn't that a song from the 80s? | 15:54 |
| * sigmavirus butchers songs for jokes | 15:54 | |
| tsymanczyk | okay yeah, if "when" were "that" i can read it the same. | 15:54 |
| sigmavirus | rosmaita: winner winner turkey dinner! | 15:54 |
| rosmaita | ok, for the bonus ... the band is "Translator" ? | 15:54 |
| sigmavirus | Uh | 15:55 |
| rosmaita | nope, modern english | 15:55 |
| sigmavirus | yeah | 15:55 |
| sigmavirus | that's what I thought (modern english) | 15:55 |
| sigmavirus | rosmaita: well you lost the stuffing | 15:55 |
| sigmavirus | but here's a turkey! | 15:55 |
| * sigmavirus throws raw turkey at rosmaita | 15:55 | |
| * rosmaita catches salmonella | 15:55 | |
| * rosmaita is no longer optimistic about anything after last week | 15:56 | |
| *** jdillaman has quit IRC | 15:56 | |
| sigmavirus | rosmaita: one of my oregon trail characters dies of salmonella | 15:56 |
| rosmaita | tsymanczyk: let me know what you think of my proposal, last comment on https://review.openstack.org/#/c/396919 | 15:58 |
| rosmaita | i mean what you think conceptually, i don't mean i'm going to ask you to do it | 15:58 |
| *** TravT has joined #openstack-glance | 15:58 | |
| rosmaita | my proposal would be to merge what's in your current patch in O-1, and then follow up with the suggested change for O-2 | 15:59 |
| *** nicolasbock has quit IRC | 16:01 | |
| *** pcaruana has joined #openstack-glance | 16:02 | |
| *** jdillaman has joined #openstack-glance | 16:05 | |
| tsymanczyk | rosmaita : remind me, please. in the current codebase if you filter on visibility=private does it only give you private visilibility / nothing shared? and visibility=shared only gives you shared / nothing private? | 16:05 |
| *** nicolasbock has joined #openstack-glance | 16:05 | |
| rosmaita | tsymanczyk: i believe that visibility=private gives you all the images you own, so if you are the image owner of shared images (that is, images that you are sharing *out* to other users), those images will show up in the response | 16:08 |
| rosmaita | but if you are an image consumer, the images shared with you will *not* show up in a visibility=private request | 16:08 |
| rosmaita | if you select visiblity=shared, you only get images that are shared with you (currently) | 16:09 |
| tsymanczyk | rosmaita : so would it be valid for me to (speciously) make the exact same opposite argument against the current codebase? where i could spend hours and hours adding members to a bunch of my images, do an image list, and be horrified to see that they're all "still private"? | 16:09 |
| rosmaita | and that are accepted | 16:09 |
| openstackgerrit | Tomislav Sukser proposed openstack/glance: IPv6 fix in Glance for malformed URLs. https://review.openstack.org/394946 | 16:10 |
| rosmaita | tsymanczyk: i think that's right, that's why some people have filed bugs about visibility==shared not showing up in an image response | 16:10 |
| tsymanczyk | rosmaita : i think your proposed solution would be a fine one (for someone else to implement :) ). the complication you point out with private visibility doesn't concern me since if a person is using that filter, they must already be aware of the changeset from the release notes or other. and if they're already aware of it and the docs were clear - then they | 16:14 |
| tsymanczyk | don't really have an excuse for confusion. (imo) | 16:14 |
| rosmaita | tsymanczyk: cool, thanks | 16:15 |
| tsymanczyk | i don't even really disagree with his issue, but it's odd to me to block a change this late in the game for a complaint that can be equally made about the current codebase. if we were making things WORSE, then okay yeah sure block it. but status quo? c'mon. | 16:15 |
| rosmaita | yes, we get some nonintuitive behavior no matter what, so my preference is to go with what makes the most sense | 16:16 |
| rosmaita | but that's kind of in the eye of the beholder | 16:16 |
| *** tshefi has quit IRC | 16:17 | |
| tsymanczyk | rosmaita : i also agree re you assessment of why we shouldn't hide "has_members" from non-shared. | 16:18 |
| rosmaita | ok, thanks | 16:19 |
| tsymanczyk | anyone want to make some sidebets about how this plays out re tomorrow's deadline :) | 16:19 |
| rosmaita | tsymanczyk: after last week, i'm not willing to make bets about anything! :) | 16:20 |
| *** belmoreira has quit IRC | 16:22 | |
| *** kfox1111_away is now known as kfox1111 | 16:28 | |
| sigmavirus | tsymanczyk: rosmaita hopefully the email I just sent will get some more eyes on this | 16:29 |
| *** e0ne has quit IRC | 16:33 | |
| *** e0ne has joined #openstack-glance | 16:34 | |
| rosmaita | sigmavirus: ty | 16:37 |
| sigmavirus | It's my job? | 16:37 |
| rosmaita | that's your answer to everything! | 16:39 |
| tsymanczyk | i appreciate the positive feedback in your email. sometimes it's easy to conflate all the -1s with nothing-but-negativity. | 16:39 |
| rosmaita | sigmavirus: anyway, i was just starting to write up a message to the ML, so you saved me some time | 16:40 |
| sigmavirus | tsymanczyk: the trick is that I don't want anyone to feel like they've done something wrong | 16:40 |
| rosmaita | tsymanczyk: yes, please try to take this positively ... erno's not criticizing your work, he's worried about the effect the change will have on consumers | 16:41 |
| sigmavirus | Right, this isn't personal | 16:41 |
| sigmavirus | I've done this to others as well | 16:41 |
| sigmavirus | It's kind of a trademark of Glance | 16:41 |
| sigmavirus | mostly because our preliminary communication is often not thorough enough and we don't know how to improve upon it | 16:41 |
| rosmaita | maybe we need a -2-with-a-smile vs -2-with-a-frown ratings | 16:41 |
| sigmavirus | Also Glance is not a priority at any company, so all of us are kind of limited in the amount of time we can squirrel away to do reviews in a timely manner | 16:42 |
| *** ckonstanski has joined #openstack-glance | 16:59 | |
| *** haplo37 has quit IRC | 17:03 | |
| *** groen692 has quit IRC | 17:03 | |
| * stevelle is trying to catch up with the latest on this | 17:06 | |
| * rosmaita is looking forward to stevelle's input | 17:07 | |
| * stevemar pokes a bunch of glancers | 17:15 | |
| stevemar | glancites? | 17:16 |
| stevemar | what do you call yourselves? | 17:16 |
| stevemar | can i get another look at https://review.openstack.org/#/c/396695/ | 17:16 |
| stevemar | its searchlight, but it seems like a lot of overlap between the two groups | 17:16 |
| *** e0ne has quit IRC | 17:21 | |
| tsymanczyk | i prefer glancistas | 17:22 |
| *** mvk has quit IRC | 17:38 | |
| jokke_ | tsymanczyk, sigmavirus, rosmaita: yes indeed, the shared by default is my only concern left | 17:41 |
| tsymanczyk | cool. thanks for confirming. | 17:43 |
| *** haplo37_ has joined #openstack-glance | 17:44 | |
| jokke_ | tsymanczyk: sorry that I did not catch that earlier ... just too busy on too many things | 17:44 |
| *** walharth_ has quit IRC | 17:47 | |
| *** lin_yang has joined #openstack-glance | 17:48 | |
| tsymanczyk | i appreciate your insightful feedback regardless. | 17:48 |
| jokke_ | tsymanczyk: to explain the translations part. The translation functions flags the changed strings to the translations team. Who in change work the correct translations and we will get automated patch providing a) removed strings from the .po-files if they are not translated in time and the strings not used anywhere in our project b) corrected translations for the new strings | 17:49 |
| jokke_ | tsymanczyk: so we should never touch the .po files manually. They all gets synced from the translation system used by the translators | 17:50 |
| tsymanczyk | perfect. i hadn't understood that, thanks for spelling it out. | 17:51 |
| jokke_ | tsymanczyk: no problem, I thought that would be the case when I saw your question about it | 17:58 |
| jokke_ | just forgot to put that on the comments in the review | 17:58 |
| openstackgerrit | Merged openstack/python-glanceclient: Move old oslo-incubator code out of openstack/common https://review.openstack.org/394780 | 17:58 |
| openstackgerrit | Dharini Chandrasekar proposed openstack/glance-specs: Return 409 if removing/replacing image locations. https://review.openstack.org/368192 | 18:00 |
| *** hemanth is now known as hemanth|lunch | 18:04 | |
| openstackgerrit | Alexander Bashmakov proposed openstack/glance: Add expand/contract commands to glance-manage CLI. https://review.openstack.org/392993 | 18:06 |
| openstackgerrit | Alexander Bashmakov proposed openstack/glance: Port Glance Migrations to Alembic https://review.openstack.org/382958 | 18:06 |
| *** e0ne has joined #openstack-glance | 18:11 | |
| *** Dong_Zhi has joined #openstack-glance | 18:22 | |
| *** e0ne has quit IRC | 18:23 | |
| *** e0ne has joined #openstack-glance | 18:37 | |
| *** harlowja has quit IRC | 18:38 | |
| tsymanczyk | stevelle: i have not formed an opinion yet, need to think further - but your 'shared' v 'sharable' idea is interesting. bloody interesting. | 18:42 |
| *** hemanth|lunch is now known as hemanth | 18:42 | |
| stevelle | tsymanczyk: :shrug: let's see where it goes. I think we might be able to build a patch on your current to get there but I'm not entirely clear how it would work yet. | 18:42 |
| *** walharthi has joined #openstack-glance | 18:57 | |
| *** rcernin has joined #openstack-glance | 19:02 | |
| *** harlowja has joined #openstack-glance | 19:06 | |
| *** walharthi has quit IRC | 19:09 | |
| *** e0ne has quit IRC | 19:31 | |
| tsymanczyk | spitballing here.... i agree that if we release it as is, SOMEBODY is going to have the bad day that jokke_ predicts ("what do you mean all my images are SHARED!? this is a disaster!"). but what if, at least for now, we were to rename the state unilaterally to "sharable" ? just the literal word "shared" means something very specific (this image IS BEING | 19:36 |
| tsymanczyk | SHARED). "sharable" implies no su ch thing. since the users already know that "private" means "potentially shared" - having the new equivalent state literally be "sharable" would be a lateral move. no? | 19:36 |
| tsymanczyk | this would not be intended to prevent or subvert any of the suggestions already posed which add value for the future. (shared v sharable, searching via membership, etc) | 19:38 |
| *** bpokorny has joined #openstack-glance | 19:49 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/glance: Updated from global requirements https://review.openstack.org/397905 | 19:49 |
| *** bpokorny has quit IRC | 19:49 | |
| *** bpokorny has joined #openstack-glance | 19:50 | |
| *** ckonstanski has quit IRC | 20:00 | |
| *** ckonstanski has joined #openstack-glance | 20:04 | |
| *** flwang1 has quit IRC | 20:06 | |
| *** TravT has quit IRC | 20:10 | |
| rosmaita | tsymanczyk: you may be onto something ... maybe instead of 'shared' we should name the visibility 'nonprivate' | 20:10 |
| rosmaita | but maybe 'sharable' is better, then instead of has_members, we could have is_shared | 20:12 |
| alex_bash | "shareable" is better | 20:12 |
| *** e0ne has joined #openstack-glance | 20:13 | |
| alex_bash | it's more intuitive for the user to understand that their images are shareable after upgrade, but not necessarily shared | 20:17 |
| rosmaita | yes, but once they are shared, the will still show as 'shareable' | 20:23 |
| rosmaita | plus, the image-list filter is visibility=shared | 20:23 |
| rosmaita | since 2.2, i think | 20:24 |
| stevelle | rosmaita and tsymanczyk that is in the range of stuff I wanted to see discussed, happy to see it might not be as terrible as I feared :) | 20:36 |
| alex_bash | then let's make shareable a separate field in the schema and remove 'shared' completely from visibility | 20:37 |
| stevelle | that might be a step too far :) | 20:37 |
| alex_bash | default to visibility=private, shareable=True | 20:37 |
| stevelle | but I'm trying to make lunch so ... | 20:37 |
| *** walharthi has joined #openstack-glance | 20:38 | |
| tsymanczyk | oh geez, i'd be against that. a separate column would be warranted if shareable had any meaning for a visibility other than private - but under the current paradigm what would it mean for a public image to have sharable=true or false? community? if the shareable flag only impacted the private visibility state, i'd argue that we'd get the same affect by | 20:52 |
| tsymanczyk | renaming 'shared' to 'private-kinda' (or whatever). | 20:52 |
| *** bpokorny has quit IRC | 20:52 | |
| tsymanczyk | rosmaita : regarding showing has-been-shared images still as 'shareable' for now doesn't concern me, as it's equivalent to what happens today (they show up as 'private' with or without members) - and just the literal word 'sharable' doesn't directly state that it isn't already being shared. the 'shared' filter ... hmm... that does make things less clean. | 20:58 |
| *** takedakn has joined #openstack-glance | 20:59 | |
| *** takedakn has quit IRC | 20:59 | |
| *** bpokorny has joined #openstack-glance | 20:59 | |
| tsymanczyk | internally though, even in the future if we did sharable v shared, i would expect that the actual visibility value in the database row would not change? internally it would be "shared" either way, the difference in display value would depend on whether or not the image has members. no? | 21:00 |
| rosmaita | tsymanczyk: that's what i'm thinking, too, but I'm not 100% sure | 21:00 |
| rosmaita | does it make sense to filter your image list by 'shareable' in addition to 'shared' ? | 21:02 |
| *** MattMan has quit IRC | 21:02 | |
| *** MattMan has joined #openstack-glance | 21:03 | |
| tsymanczyk | i can certainly see a case where a person would want all their images that are currently-being-shared. other than contrived examples, i'm not sure why i'd want to see "all my images that are sharable but not-currently"... | 21:03 |
| *** e0ne has quit IRC | 21:04 | |
| rosmaita | on your current implementation, if I want to see the shared images that I own, what query do i make? | 21:04 |
| tsymanczyk | but going back to your idea from this morning, if the visibility value were 'sharable' - period - then we added the "has_members" filter, that'd address both cases with a single visibility. just alias shared / sharable to each other filterwise maybe | 21:04 |
| tsymanczyk | visibility=shared&owner=pretty_me | 21:04 |
| rosmaita | ok | 21:05 |
| tsymanczyk | i think i like your idea more and more | 21:06 |
| tsymanczyk | visibility=sharable&owner=pretty_me&has_members=true | 21:06 |
| tsymanczyk | what's wrong with that? | 21:06 |
| openstackgerrit | Dharini Chandrasekar proposed openstack/glance-specs: Return 409 if removing/replacing image locations. https://review.openstack.org/388950 | 21:06 |
| rosmaita | well, the only thing is that we already have visibility=shared is already defined in the API | 21:06 |
| tsymanczyk | have a little piece of code that silently interprests shared as sharable | 21:07 |
| tsymanczyk | let them specify either one | 21:07 |
| tsymanczyk | we've done grosser things for backwards compat | 21:07 |
| rosmaita | the problem is, if we call it 'shared', then people wonder, is it already available to other people, and if we call it 'shareable', the owner wonders the same thing | 21:08 |
| *** ccesario has joined #openstack-glance | 21:09 | |
| tsymanczyk | are you sure about that? the users already live in a world where 'private' doesn't mean PRIVATE. if anything i'd think 'shareable' is less ambiguous. | 21:09 |
| tsymanczyk | i'm totally onboard with jokke's issue though. 'shared' would scare people. | 21:10 |
| rosmaita | yeah | 21:12 |
| rosmaita | so you like 'shareable' as the visibility identifier, plus adding has_members or member_count to indicate whether it's actually available to other users or not | 21:12 |
| tsymanczyk | of all the options i'm aware of, that seems to me to be the cleanest. yes. | 21:14 |
| tsymanczyk | plus, as you've already pointed out, if member_count is included as a field in all image-detail-calls, then (as you already implied earlier?) it addresses the nagging potential issue where non-shared images still maintain their user lists / but non-shared they're inaccessible. so a person with a currently private image when switching to 'shared' has no way | 21:16 |
| tsymanczyk | to know ahead of time (short of db inspection) how many, if any, members they're going to START with. | 21:16 |
| tsymanczyk | hm.... it's kind of gross though. | 21:19 |
| * alex_bash is worried about the complexity of the db trigger growing exponentially | 21:19 | |
| tsymanczyk | if we SHOW the user that this private image has a non-zero member count, then we're kind of being .... arbitrary? by not allowing them to edit it. | 21:20 |
| tsymanczyk | otherwise it's "here is the situation - you can't DO anything about it. but here it is." | 21:20 |
| rosmaita | well, you can do something, but you have to change the visibility first | 21:21 |
| rosmaita | problem is that you can't remove members unless you make the image in viz shared | 21:21 |
| rosmaita | which opens the image to someone you don't want to have it until you remove them | 21:22 |
| tsymanczyk | i'm worried we're about to step into the rabbit hole that we explicitly jumped over for the original spec. the "state-transition-affect-on-member-list" section that was deemed complex enough to warrant its own spec later on. | 21:22 |
| rosmaita | yeah | 21:23 |
| tsymanczyk | would it be so bad if we let people edit the membership list regardless of visibility? things got complicated back then because we were expecting to use the same table and mechanism for community images bookmarking. | 21:23 |
| tsymanczyk | if it were CLEAR that "the membership list of any private/public/community image can be editted but has no effect at all outside of the shared visibility" - then ... maybe? | 21:24 |
| rosmaita | i think you are right that this is a rabbit hole we should avoid ATM | 21:24 |
| tsymanczyk | done | 21:25 |
| *** hemanth is now known as hemanth|away | 21:26 | |
| tsymanczyk | okay, realistically, speaking only about the one very real current objection. | 21:27 |
| tsymanczyk | switching from the word 'shared' to the word 'sharable', and internally aliasing shared/sharable to be the same... what is lacking in that solution? it addresses the immediate problem, and leaves the future plans wide open. | 21:27 |
| rosmaita | suppose we go with viz='shareable' and has_members=<bool> on the image object | 21:27 |
| tsymanczyk | (to be the same visibility filter wise) | 21:27 |
| rosmaita | right | 21:28 |
| tsymanczyk | even if the operator panics, not understanding the literal word, they're two calls away from seeing that nothing has changed. (has_members=false, has_members=true). | 21:31 |
| tsymanczyk | that's not awful | 21:31 |
| *** pcaruana has quit IRC | 21:32 | |
| stevelle | I'm pretty happy with 'shareable' as the visibility identifier, plus (later) adding has_members or member_count. I'm not adverse to allowing DELETE of members on private images also. | 21:38 |
| tsymanczyk | if this is the path forward, i can take care of pushing a new patchset immediately. | 21:41 |
| stevelle | if we can get feedback from jokke_ first that would be helpful. | 21:42 |
| *** walharthi has quit IRC | 21:43 | |
| *** flwang1 has joined #openstack-glance | 21:49 | |
| tsymanczyk | not that this should affect anyone's decisions or process, but after O-1 is cut, i won't be nearly as available or responsive. i've argued to my company "for this time now" so i can make a clean break for the future rather than devoting ~X% a week to polishing this forever. | 21:52 |
| tsymanczyk | on the assumption that being extremely-responsive now would give the best chance of merging into o-1 | 21:53 |
| *** walharthi has joined #openstack-glance | 21:54 | |
| *** harlowja has quit IRC | 21:55 | |
| stevelle | it's really helpful that you have been able to continue being responsive still, tsymanczyk. | 21:59 |
| rosmaita | tsymanczyk: i second what stevelle said | 22:03 |
| tsymanczyk | thank you both - i'm happy to do it, i want to see this succeed. i'm just saying the real world is about to cave in on me. | 22:04 |
| *** harlowja has joined #openstack-glance | 22:13 | |
| *** mvk has joined #openstack-glance | 22:14 | |
| *** catinthe_ has joined #openstack-glance | 22:18 | |
| *** catintheroof has quit IRC | 22:21 | |
| *** walharthi has quit IRC | 22:27 | |
| *** catintheroof has joined #openstack-glance | 22:35 | |
| *** catinthe_ has quit IRC | 22:36 | |
| *** edmondsw has quit IRC | 22:36 | |
| *** rcernin has quit IRC | 22:42 | |
| *** rrecio_ has quit IRC | 23:05 | |
| *** walharthi has joined #openstack-glance | 23:08 | |
| *** ducttape_ has quit IRC | 23:27 | |
| *** ckonstanski has quit IRC | 23:39 | |
| *** ckonstanski has joined #openstack-glance | 23:40 | |
| *** lujinluo has joined #openstack-glance | 23:46 | |
| *** catintheroof has quit IRC | 23:52 | |
| *** catintheroof has joined #openstack-glance | 23:54 | |
| *** bpokorny has quit IRC | 23:55 | |
| *** catintheroof has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!