| *** salv-orlando has quit IRC | 00:03 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 02:12 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 02:15 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 02:40 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 02:52 | |
| *** HenryG has quit IRC | 03:51 | |
| *** alex_xu has joined #openstack-api | 05:24 | |
| *** HenryG has joined #openstack-api | 05:38 | |
| openstackgerrit | Alex Xu proposed openstack/api-wg: Clarify the return code when server have hard-code length limit https://review.openstack.org/181784 | 05:57 |
|---|---|---|
| *** fzdarsky has joined #openstack-api | 07:09 | |
| *** fzdarsky has quit IRC | 07:17 | |
| *** gilliard is now known as gilllliard | 07:26 | |
| *** fzdarsky has joined #openstack-api | 07:30 | |
| *** woodster_ has quit IRC | 08:10 | |
| *** e0ne has joined #openstack-api | 09:55 | |
| *** e0ne is now known as e0ne_ | 09:56 | |
| openstackgerrit | Sean Dague proposed openstack/api-wg: http guideline expansion: background https://review.openstack.org/181931 | 09:59 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add section describing 501 common mistake https://review.openstack.org/183456 | 09:59 |
| *** e0ne_ is now known as e0ne | 10:01 | |
| *** cdent has joined #openstack-api | 10:06 | |
| cdent | HTTP is such a fine source of bikeshedding material | 10:07 |
| sdague | heh | 10:11 |
| sdague | I was wondering if you were going to be online today | 10:11 |
| sdague | so... the 501 thing is why I still don't want "go to the rfc" | 10:11 |
| sdague | as step 2 | 10:11 |
| sdague | that's how we got there | 10:12 |
| sdague | I hopefully updated the wording on https://review.openstack.org/181931 to be ok with you | 10:12 |
| sdague | also - https://review.openstack.org/183456 | 10:12 |
| sdague | and you can read the initial conversation that got us there | 10:12 |
| cdent | the final paragraph with “appropriate references to rfcs” is great and satisfies me | 10:12 |
| cdent | I’m happy with it | 10:12 |
| sdague | cool | 10:12 |
| cdent | yeah, I was just on my way to that list thread | 10:13 |
| cdent | I _should_ be packing and otherwise preparing myself (my journey starts tomorrow early afternoon but I won’t be there until sunday evening) but I can’t seem to stay away... | 10:14 |
| sdague | so, chris ferris and I argued against what was decided, but he had 0 pull, as he was a random architect at IBM, and my pull was very limitted. I eventually caved because I was arguing against one of the highest visibility devs at the time | 10:14 |
| sdague | heh | 10:14 |
| sdague | my journey also starts tomorrow afternoon, but I need the rest of my house to wake up before I start figuring out packing | 10:15 |
| cdent | If I had thought in advance I would own five pairs of exactly the same jeans and ten black t shirts: 5 short sleeved, 5 long | 10:16 |
| cdent | and then I’d be done | 10:16 |
| cdent | but I guess I’m not that kind of person | 10:16 |
| cdent | I might be forced to wear this on at leat one day: http://www.zazzle.com/business_customized_t_shirt-235989706520424324 | 10:16 |
| cdent | sdague: thanks for point to that bit of history. fascinating in the way it sort of illuminates some of our modern luminaries... | 10:21 |
| sdague | heh | 10:24 |
| sdague | both on the shirt, and on the comments :) | 10:25 |
| *** e0ne is now known as e0ne_ | 10:31 | |
| *** e0ne_ is now known as e0ne | 10:33 | |
| *** cdent has quit IRC | 10:52 | |
| *** e0ne is now known as e0ne_ | 11:26 | |
| *** e0ne_ is now known as e0ne | 11:48 | |
| gilllliard | oh well, | 11:56 |
| gilllliard | interesting reading on the old ML, thanks sdague | 11:56 |
| gilllliard | Nova API meeting in 3 minutes, btw | 11:57 |
| gilllliard | #openstack-meeting | 11:57 |
| sdague | gilllliard: yeh, I figured that historical context would be useful | 11:57 |
| sdague | I'm surprised it wasn't cancelled this week | 11:58 |
| sdague | given all the inflight people | 11:58 |
| sdague | I've got to get back to packing, so will not be there | 11:58 |
| gilllliard | sdague: it wasn't cancelled, and there's some agenda, so I'll pop along and see if anyone else is there. | 11:59 |
| *** gilllliard is now known as gilliard_en_rout | 12:09 | |
| *** e0ne is now known as e0ne_ | 12:22 | |
| *** e0ne_ is now known as e0ne | 12:26 | |
| *** annegentle has joined #openstack-api | 12:48 | |
| *** e0ne is now known as e0ne_ | 12:54 | |
| *** e0ne_ is now known as e0ne | 12:56 | |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add section describing 501 common mistake https://review.openstack.org/183456 | 13:23 |
| *** woodster_ has joined #openstack-api | 13:41 | |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add subsection around caching behavior and http https://review.openstack.org/183523 | 13:45 |
| sdague | ryansb: so I coppied the format of other links in there, what's the right link syntax I should use? I seriously always mix up all the link semantics in my head | 13:53 |
| ryansb | sdague: `link title<http://link.com/foo>`_ | 13:54 |
| sdague | ryansb: ok, great, I'll fix this last patch | 13:54 |
| ryansb | (rst link syntax makes me very sad compared to markdown) | 13:54 |
| ryansb | thanks :) | 13:55 |
| *** cdent has joined #openstack-api | 13:55 | |
| ryansb | you can also do | 13:55 |
| ryansb | Link_ | 13:55 |
| ryansb | .. _Link: http://foo.bar/baz | 13:55 |
| ryansb | but that's a bit harder to follow | 13:55 |
| sdague | yeh, that seems less source readable | 13:56 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add subsection around caching behavior and http https://review.openstack.org/183523 | 13:56 |
| sdague | ok, I guess I should fix the whole stack, eh? | 13:57 |
| *** annegentle has quit IRC | 13:59 | |
| openstackgerrit | Sean Dague proposed openstack/api-wg: http guideline expansion: background https://review.openstack.org/181931 | 13:59 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add subsection around caching behavior and http https://review.openstack.org/183523 | 13:59 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add section describing 501 common mistake https://review.openstack.org/183456 | 13:59 |
| sdague | ryansb: ok, that fixes the links in the whole stack | 14:00 |
| ryansb | sdague: :) | 14:00 |
| elmiko | i gotta move faster, my drafts keep getting over-written ;) | 14:01 |
| elmiko | you're too fast sdague! | 14:01 |
| sdague | what's the approval / merge process look like for the group. As I'm relatively new in contributing, mostly spurred on by cdent and jaypipes discussions over the last couple of weeks. | 14:01 |
| sdague | elmiko: :) | 14:01 |
| sdague | I also take feedback over irc | 14:01 |
| sdague | :) | 14:01 |
| elmiko | we have usually been pretty conservative with our merge process. allow time for everyone to take a look and comment. | 14:02 |
| sdague | ok, no prob. | 14:02 |
| elmiko | usually things get merged with like +8 or more responding | 14:02 |
| cdent | The merging process appears to be complete magic | 14:02 |
| sdague | hopefully most of this is pretty ack kind of stuff, except for that 501 discussion | 14:02 |
| cdent | people talk, have coffee, and then after a while stuff merges | 14:03 |
| elmiko | yea, what cdent says is pretty accurate | 14:03 |
| sdague | cdent: I have a new subsection up that you will probably appreciate | 14:03 |
| sdague | ok, there is a little girl demanding attention here, so away for a bit | 14:03 |
| elmiko | cdent: i don't think it's complete magic, but there is some mystery | 14:05 |
| elmiko | ;P | 14:05 |
| elmiko | the way i see it, patches are up for long enough that everyone could get a look and comment. | 14:05 |
| cdent | I know that there is some logic behind it, it’s not obvious. And I think that’s fine. It seems that things pootle along for a while and then when various things signal consensus a merge happens. | 14:06 |
| elmiko | since many folks don't check the api-wg reviews that frequently it can take time to build consensus | 14:06 |
| elmiko | yea, i think that's accurate | 14:06 |
| cdent | It works! \o/ | 14:08 |
| *** salv-orlando has joined #openstack-api | 14:08 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:24 | |
| etoews | sdague: here's what we have in terms of formal process http://specs.openstack.org/openstack/api-wg/process.html | 14:24 |
| etoews | for merging that is | 14:25 |
| elmiko | ooh, good link! | 14:25 |
| etoews | the more informal goes something like this: | 14:28 |
| etoews | 1. discuss within the wg until consensus is reached within the wg | 14:28 |
| etoews | 2. expand to cross-project liaisons (this is where it gets a bit fuzzy) | 14:28 |
| etoews | 3. bring up consensized guidelines at a cross-project meeting | 14:28 |
| etoews | 4. "freeze" the guideline with a -1 workflow for 1 week while waiting for feedback from CPLs | 14:28 |
| etoews | 5. merge after 1 week on lazy consensus or rework as necessary | 14:28 |
| cdent | consensized is a nice word | 14:29 |
| etoews | we really only arrived at that process a couple of weeks ago. probably why it appears to be magic. | 14:29 |
| etoews | i'm definitely happy to accelerate the process | 14:30 |
| elmiko | etoews: how strongly should we follow #4, is that a hard -1 after the cross-project discussion? | 14:30 |
| etoews | the -1 workflow is just so it doesn't get merged "accidentally" | 14:30 |
| elmiko | got it | 14:30 |
| etoews | sigmavirus24 and someone else suggested it. | 14:30 |
| sigmavirus24 | I am to blame | 14:31 |
| sigmavirus24 | I'm sorry | 14:31 |
| sigmavirus24 | And you're welcome | 14:31 |
| etoews | :) | 14:31 |
| elmiko | it's a good idea | 14:31 |
| *** annegentle has joined #openstack-api | 14:31 | |
| etoews | this is good fodder for discussion at the summit | 14:31 |
| etoews | we could take one of the working sessions to update http://specs.openstack.org/openstack/api-wg/process.html with this info if all are agreeable | 14:32 |
| elmiko | that could be cool | 14:32 |
| etoews | added it to the etherpad https://etherpad.openstack.org/p/liberty-cross-project-api-wg | 14:34 |
| ryansb | I like that merge process (though it isn't the fastest) since our specs rely so hard on consensus/acceptance from project devs | 14:42 |
| elmiko | agreed, i think it makes sense | 14:44 |
| *** salv-orlando has quit IRC | 14:46 | |
| openstackgerrit | Merged openstack/api-wg: Change RFC2616 links to RFCs 723[0-5] https://review.openstack.org/177445 | 14:59 |
| *** cdent_ has joined #openstack-api | 15:09 | |
| *** cdent has quit IRC | 15:09 | |
| *** cdent_ is now known as cdent | 15:09 | |
| cdent | etoews: “I've seen it happen with 4xx status codes”. Where was that? That’s a) surprising b) saddening. | 15:27 |
| etoews | i think it was heat | 15:28 |
| etoews | but working off devstack so it may have been in debug mode | 15:28 |
| cdent | even so, still weird | 15:28 |
| etoews | i also swear i've seen it before but can't quite recall exactly where | 15:29 |
| cdent | but then again, I recently discovered that WSME was logging 4xx as WARN not DEBUG | 15:29 |
| cdent | which struck me as completely surprising | 15:29 |
| etoews | danger. the user is doing something wrong. | 15:30 |
| cdent | panic. now. | 15:30 |
| ryansb | whoa, users make mistakes? Hold up. | 15:42 |
| sdague | etoews: yeh, I'm unlikely to make the session because of all the cross neutron / nova network things. But will try to connect. | 15:42 |
| etoews | sdague: sounds good. we're also having follow up working sessions on wed and thurs. see the bottom of https://etherpad.openstack.org/p/liberty-cross-project-api-wg | 15:45 |
| sdague | yeh... well... being involved in nova means my design summit week is mostly predetermined. Perhaps drinks at some point. | 15:49 |
| elmiko | hehe | 15:49 |
| etoews | at least the keeps it simple for you :) | 15:57 |
| etoews | so i'm looking at https://review.openstack.org/#/c/181784/6/guidelines/http.rst | 15:57 |
| etoews | to me this sounds like basic input validation stuff, just return a 400. | 15:58 |
| etoews | does it really need a guideline? seems like this is covered by rfcs. | 15:58 |
| cdent | aye, there’s the rub | 15:58 |
| elmiko | good point, but didn't this specific example come up at some point? | 15:58 |
| cdent | the specific example did come up and the getting this particular submission to come round to 400 was a bit of struggle | 15:59 |
| elmiko | imo, this is ripe material for separating the response codes into a separate guidelines and to make use of the "Examples" section for clarity. | 16:00 |
| cdent | etoews: originally it was 403, and I assume the introduction of the 403 text, now that it has been changed to 400, is to clarify that doubt | 16:01 |
| cdent | this goes back, as we often do, to the question of how much are we supposed to be recapitulating the rfcs and/or expecting to already know them | 16:02 |
| etoews | i'm fine recapitulating the rfcs when there's some ambiguity or there a reason openstack does things some particular way. | 16:04 |
| etoews | but this is input validation | 16:04 |
| cdent | apparently for this person there was ambiguity | 16:04 |
| etoews | i would hope that returning a 400 doesn't need to be recapitulated. | 16:04 |
| *** annegentle has quit IRC | 16:04 | |
| etoews | right | 16:04 |
| cdent | I would hope that too, but the evidence suggests otherwise | 16:04 |
| etoews | i might need to ... calibrate ... my perspective on this. | 16:05 |
| cdent | yes, I’ve had to do a _lot_ of calibration | 16:07 |
| cdent | And realize that the main win we’re ending up with in this stuff is an educational resource | 16:07 |
| elmiko | +1 | 16:07 |
| etoews | that's exactly where my thinking was going too | 16:07 |
| elmiko | i find it to be helpful | 16:08 |
| elmiko | but i'm not sure how much of the rfc's we are restating | 16:08 |
| etoews | perhaps there's a larger educational role for the api wg than i anticipated. | 16:08 |
| cdent | I was going through Jay’s slides and realized that his goals for the wg were a lot more about api consistency, which is good, but it seems a distance off from just doing http okay | 16:08 |
| elmiko | to build consistency though, especially with a variable developer pool, don't we have to educate a little bit? | 16:09 |
| elmiko | sorry to question and run, but i gtg. back later | 16:09 |
| cdent | elmiko: yes, exactly | 16:09 |
| etoews | okay. this really helps me a lot. | 16:10 |
| cdent | From what I’ve experienced a vast number of the people contributing to openstack are not hugely experience in either http or python. | 16:10 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: http guideline expansion: background https://review.openstack.org/181931 | 16:10 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add subsection around caching behavior and http https://review.openstack.org/183523 | 16:10 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add section describing 501 common mistake https://review.openstack.org/183456 | 16:10 |
| cdent | And that’s great in the sense that we’ve got a big community | 16:10 |
| cdent | but it does imply a requirement to be educational | 16:10 |
| sdague | ok, rebased after the rfc update bits, and fixing some of the latest comments and links | 16:11 |
| sdague | cdent: I would 100% agree with you. | 16:11 |
| sdague | etoews: so... I would say 80+% of the folks that are implementing the API side of openstack services, this project was the first time they did anything at the HTTP level | 16:12 |
| sdague | so... recapitulating basic http stuff is needed | 16:13 |
| sdague | otherwise, you get the 501 situation | 16:13 |
| cdent | ™ | 16:13 |
| sdague | I think looking at how that conversation happened historically, where we ended up, and what it's going to take to reverse it is really educational in grasping where the basic knowledge levels started with | 16:14 |
| etoews | sdague: that's very good to know. i added a note about the "educational aspect" to the etherpad and i think i'll update my blog post too. | 16:14 |
| *** e0ne is now known as e0ne_ | 16:15 | |
| sdague | cdent: so I'll at one more thing to your first time list | 16:16 |
| sdague | for a vast number of people contributing to openstack, this is the first open source project they've ever contributed to | 16:16 |
| cdent | yeah | 16:16 |
| cdent | I had a realization the other day: I’ve never worked on not open source | 16:17 |
| sdague | etoews: and, honestly, I think getting the logging guidelines written down last year was a really interesting educational experience of "most people just had no idea things they were doing were whacky, and were happy to change as long as someone told them why and how" | 16:18 |
| sdague | https://blogs.gnome.org/markmc/2014/06/06/an-ideal-openstack-developer/ gets to the heart of it kind of nicely | 16:19 |
| sdague | there is a ton of stuff one really *should* know to contribute to openstack, to the point that no one actually does | 16:19 |
| cdent | I saw that talk (in London) | 16:19 |
| *** e0ne_ is now known as e0ne | 16:20 | |
| sdague | it would have been fun to see :) | 16:20 |
| etoews | my perspective has been calibrated :) | 16:20 |
| cdent | was that that grinding noise? | 16:20 |
| etoews | smoke is pouring out of my ears. | 16:21 |
| sdague | heh | 16:21 |
| sdague | it's fine, as long as you think of it as a long game, things get written down now, and 2 or 3 years later, it's better | 16:21 |
| * cdent notes he still hasn’t started packing | 16:23 | |
| sdague | heh, I did accomplish that | 16:24 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: http guideline expansion: background https://review.openstack.org/181931 | 16:27 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add subsection around caching behavior and http https://review.openstack.org/183523 | 16:27 |
| openstackgerrit | Sean Dague proposed openstack/api-wg: add section describing 501 common mistake https://review.openstack.org/183456 | 16:27 |
| sdague | that should finally fix those sphinx links | 16:27 |
| sdague | man, that is a gorpy and fragile syntax | 16:27 |
| * cdent is not a fan | 16:28 | |
| * cdent communes with his suitcase | 16:34 | |
| openstackgerrit | Jay Pipes proposed openstack/api-wg: Add guidance on 500 Internal Server Error https://review.openstack.org/179365 | 16:35 |
| *** e0ne has quit IRC | 16:37 | |
| openstackgerrit | Jay Pipes proposed openstack/api-wg: Should not return server-side tracebacks https://review.openstack.org/183599 | 16:40 |
| openstackgerrit | Jay Pipes proposed openstack/api-wg: Adds clarifications on state-conflicting requests https://review.openstack.org/180094 | 16:40 |
| openstackgerrit | Jay Pipes proposed openstack/api-wg: Adds clarifaction on when to use 409 https://review.openstack.org/179386 | 16:41 |
| *** subscope has joined #openstack-api | 16:47 | |
| *** fzdarsky has quit IRC | 17:00 | |
| *** annegentle has joined #openstack-api | 17:05 | |
| *** annegentle has quit IRC | 17:10 | |
| *** e0ne has joined #openstack-api | 17:15 | |
| *** annegentle has joined #openstack-api | 17:35 | |
| *** annegentle has quit IRC | 17:40 | |
| *** subscope has quit IRC | 17:44 | |
| *** subscope has joined #openstack-api | 17:49 | |
| *** annegentle has joined #openstack-api | 18:11 | |
| *** openstackgerrit has quit IRC | 18:22 | |
| *** openstackgerrit has joined #openstack-api | 18:23 | |
| *** cdent has quit IRC | 18:23 | |
| sdague | etoews: so for stuff like the comment here - https://review.openstack.org/#/c/183456/ how would you feel about setting ground rules that people should just fix and upload changes like that? Instead of adding review cycle delay. | 19:18 |
| sdague | we often do that in nova, just fix the typo, then stick our +2 on it | 19:18 |
| elmiko | doc team does that too, iirc | 19:21 |
| sdague | having realized that the etherpad on http has just lived in this channel, I just sent out an email about it. | 19:24 |
| sdague | see if people are up for grabbing a bullet or two before getting on a plane. | 19:25 |
| elmiko | awesome! | 19:28 |
| elmiko | sdague: so, we have the 2xx codes stuff in there. did you envision it changing? | 19:32 |
| *** annegentle has quit IRC | 19:33 | |
| elmiko | maybe it should be formatted more as natural text than bullet points | 19:34 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 19:38 | |
| elmiko | sdague: (looking at the etherpad) i'm a little confused about the "never use 5xx" comment. shouldn't it be used if the server has an error? | 19:40 |
| *** subscope has quit IRC | 19:54 | |
| sdague | elmiko: the applications shouldn't intentionally create a 5xx | 19:56 |
| sdague | so, the 2xx content is currently totally reasonable, we may want to think about clarity later, I was mostly going to leave as is | 19:56 |
| elmiko | sdague: ok, so only the framework should be returning a 5xx. like if an exception gets thrown. | 19:56 |
| sdague | yeh | 19:56 |
| elmiko | ok. | 19:57 |
| elmiko | i'm not quite confident enough in my advice to tackle one of these, but i'm trying to understand more. | 19:57 |
| *** fzdarsky has joined #openstack-api | 19:59 | |
| sdague | ok, no problem | 19:59 |
| elmiko | i'll try and take a stab at the 5xx case, i think that makes sense to me | 20:00 |
| sdague | maybe it will spur cdent into snagging some | 20:00 |
| sdague | elmiko: cool, thank you | 20:00 |
| elmiko | sdague: do you think it's better to create a new patch from master or depend on your chain? | 20:02 |
| sdague | I'm not sure it matters | 20:02 |
| elmiko | k | 20:02 |
| elmiko | thanks for the advice =) | 20:02 |
| sdague | there will be some merge conflicts, but they will be easy to resolve | 20:02 |
| elmiko | for sure | 20:02 |
| elmiko | and then for 1xx, we should just never be generating them, correct? | 20:03 |
| sdague | correct | 20:05 |
| elmiko | cool | 20:05 |
| sdague | etoews: https://review.openstack.org/#/c/131725 - seems stalled out? | 20:11 |
| elmiko | maybe we should combine it and the dependency into a new CR using the template? | 20:14 |
| etoews | sdague: looking... | 20:14 |
| etoews | sdague: that one seems to be superseded by https://review.openstack.org/#/c/131726/ | 20:17 |
| etoews | odd to do those as 2... | 20:17 |
| elmiko | yea | 20:17 |
| etoews | we do want to say something about api documentation | 20:18 |
| etoews | it came up with annegentle a week or two back | 20:19 |
| etoews | and we agreed that documentation was in scope for creating guidelines around | 20:21 |
| etoews | however, i know for a fact that the original author of those patch sets, Sam Harwell, is no longer working on anything openstack related. | 20:21 |
| etoews | i'm pretty sure we won't be seeing any updates from him | 20:22 |
| etoews | so either we patch over them or abandon the patch sets altogether | 20:22 |
| etoews | i'd like to get anne's opinion before we make a decision there | 20:24 |
| etoews | but i suspect she won't have a chance to render an opinion before the summit | 20:24 |
| elmiko | hmm, reading through that second CR i think this may be a large can of worms... | 20:24 |
| sdague | yeh, it seems just orthoginal at this point realistically | 20:25 |
| *** fzdarsky has quit IRC | 20:30 | |
| openstackgerrit | Michael McCune proposed openstack/api-wg: Adding 1xx guidance https://review.openstack.org/183694 | 20:34 |
| *** annegentle has joined #openstack-api | 20:44 | |
| annegentle | hey | 20:45 |
| elmiko | hi | 20:46 |
| openstackgerrit | Michael McCune proposed openstack/api-wg: Adding 5xx guidance https://review.openstack.org/183698 | 20:46 |
| openstackgerrit | Michael McCune proposed openstack/api-wg: Adding 1xx guidance https://review.openstack.org/183694 | 20:48 |
| *** annegentle has quit IRC | 21:49 | |
| *** annegentle has joined #openstack-api | 21:51 | |
| *** annegentle has quit IRC | 22:51 | |
| *** elmiko is now known as _elmiko | 22:57 | |
| *** terrylhowe has quit IRC | 23:03 | |
| *** e0ne has quit IRC | 23:04 | |
| *** annegentle has joined #openstack-api | 23:51 | |
| *** annegentle has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!