20:00:27 <david-lyle> #startmeeting Horizon
20:00:27 <openstack> Meeting started Wed Dec  2 20:00:27 2015 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:00:29 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:00:32 <openstack> The meeting name has been set to 'horizon'
20:00:35 <david-lyle> any horizoneers?
20:00:45 <mrunge> o/
20:00:50 <matt-borland> o/
20:00:50 <rhagarty> o/
20:01:02 <tsufiev> o/
20:01:11 <ducttape_> o/
20:02:11 <bpokorny> Hi
20:03:08 <david-lyle> Let's start with M-1
20:03:49 <david-lyle> M-1 is due UTC tomorrow
20:04:11 <david-lyle> so I'm attempting to read up on the new tagging procedure and make that happen
20:04:21 <david-lyle> we should have an M-1 tomorrow
20:04:54 <david-lyle> next status
20:05:02 <david-lyle> we have two versions of reality
20:05:11 <tsufiev> just 2?
20:05:13 <david-lyle> #link https://launchpad.net/horizon/+milestone/mitaka-1
20:05:23 <david-lyle> and
20:05:26 <david-lyle> #link https://etherpad.openstack.org/p/mitaka-horizon-priorities
20:06:05 <r1chardj0n3s> o/
20:07:02 <david-lyle> launchpad is a view of what merged and etherpad is more for tracking
20:07:06 <david-lyle> I guess
20:08:00 <david-lyle> since we're at M-1, I'd like to look at the priorities real quick
20:08:23 <david-lyle> documentation is still inflight
20:08:36 <david-lyle> at least one patch to review and some have merged
20:08:46 <david-lyle> if there are more please update the list
20:09:24 <david-lyle> The angular work has a huge list of patches
20:09:44 <david-lyle> with some merging, probably not as quickly as some would hope
20:09:44 <matt-borland> yep
20:09:51 <matt-borland> it's going pretty well
20:09:58 <david-lyle> looks like progress to me
20:10:01 <matt-borland> we've merged 20 angular patches since the summit
20:10:28 <david-lyle> the critical part of theming (documentation) was merged
20:10:43 <david-lyle> the dependencies dance continues
20:10:56 <david-lyle> latest suggestion is setuptools_scm
20:11:09 <david-lyle> which I have to read up on
20:11:15 <r1chardj0n3s> yeah, the xstatic story continues
20:11:28 <david-lyle> act 17 in 403 act play
20:11:46 <r1chardj0n3s> I'm working on a thing to publish xstatic using setuptools_scm instead of pbr, but we still need to get that tool through approval to be used
20:11:58 <r1chardj0n3s> since it needs to be pre-loaded into the build system alongside pbr
20:12:23 <david-lyle> so simple
20:12:25 <r1chardj0n3s> but that should hopefully be a lot easier than changing pbr
20:12:37 <david-lyle> ok, that's promising
20:12:37 <r1chardj0n3s> I like your optimism!
20:13:06 <tsufiev> r1chardj0n3s, does setuptools_scm have more active developers than pbr :)?
20:13:26 <r1chardj0n3s> tsufiev: the benefit is we don't need to alter setuptools_scm
20:13:38 <r1chardj0n3s> (or, hopefully, type it very much, since that's a lot to type ;-)
20:14:14 <david-lyle> sahara-dashboard now contains all the horizon sahara content \o/ (other than integration tests)
20:14:45 <david-lyle> I need to update patch to remove to get rid of integration tests
20:14:53 <david-lyle> then it should pass python34 job
20:15:27 <david-lyle> any other high items people wanted to discuss?
20:15:50 <tsufiev> david-lyle, shall we wait until sahara-dashboard merges the integraiton tests before removing them from Horizon repo?
20:16:07 <david-lyle> tsufiev: working on that, but yeah
20:16:26 <tsufiev> david-lyle, ok, I'll prioritize this as well for myself
20:16:27 <david-lyle> hence no update to the removal patch
20:16:50 <david-lyle> some go_to method magic is slowing me down now
20:17:34 <david-lyle> ok, moving on
20:17:53 <bpokorny> I think this one is ready to go:  https://review.openstack.org/#/c/196328/
20:17:59 <bpokorny> Unless anyone still has concerns on it.
20:18:22 <tsufiev> david-lyle, yep, it's magic indeed :/
20:18:24 <david-lyle> bpokorny: I'm close to pulling the trigger on that one
20:18:39 <bpokorny> Sounds good.  Thanks, david-lyle.
20:18:46 <david-lyle> then we'll release d-o-a
20:18:58 <bpokorny> Great
20:19:33 <david-lyle> Remaining general item is Midcycle
20:19:53 <david-lyle> I failed to post a survey for dates and locations
20:20:33 <david-lyle> anyone with extra time and a driving interest that wants to take that on? I'm happy to work with them, or I'll try to get it up this week
20:21:41 <mrunge> david-lyle, just the post to mailing list, or is there more to do?
20:21:53 <david-lyle> creating the survey
20:22:10 <mrunge> so do we already have proposals?
20:22:24 <Piet> I don't mind creating a survey for ya'll
20:22:37 <david-lyle> mrunge: we have several willing sites, just need dates
20:22:52 <Piet> What are the sites?
20:22:56 <mrunge> Piet already offered, but I can do too
20:23:15 <david-lyle> or if a europe midcycle is on the table, we'd need to find a site for that
20:23:39 <Piet> Intel could potentially host in London
20:24:00 <tsufiev> interesting...
20:24:15 <david-lyle> Piet, I'll work with you offline
20:24:16 <Piet> They already offered to host the UX mid-cycle if you want to dovetail
20:24:27 <r1chardj0n3s> I could also help with the survey side of things if extra hands are needed
20:24:48 <david-lyle> I think more hands than mine are needed :)
20:25:14 <r1chardj0n3s> so first we need a call for locations, just so we're clear who is offering to host
20:25:22 <Piet> If you have locations/dates it shouldn't take me more than thirty minutes in google forms
20:25:54 <david-lyle> r1chardj0n3s, Piet: I have a list of several locations
20:26:02 <david-lyle> the london bit is new thought
20:26:05 <david-lyle> *though
20:26:09 <Piet> Another potential would be to host at IBM's design center in Austin
20:26:37 <david-lyle> Let's take it offline, maybe after this meeting
20:26:41 <Piet> k
20:26:49 <r1chardj0n3s> k
20:27:42 <david-lyle> Last general items, I've proposed a couple of new horizon cores on the mailing list.
20:27:59 <r1chardj0n3s> thanks, david-lyle :-)
20:28:06 <tsufiev> david-lyle, thanks :)
20:28:09 <Piet> I have one other thing
20:28:28 <david-lyle> The agenda for today was blank, so
20:28:35 <david-lyle> #topic general discussion
20:28:41 <david-lyle> go Piet
20:28:50 <Piet> UX Cores please review and - 2 / +2  https://invis.io/VU550E82R
20:29:08 <Piet> Uhhhh one other thing
20:29:31 <Piet> As you know, I will be moving to Intel on Monday
20:29:47 <r1chardj0n3s> (I didn't know, congrats Piet!)
20:29:57 <mrunge> congrats Piet
20:30:09 <Piet> In the meantime, pkruithofjr   <@> gmail.com
20:30:10 <tsufiev> Piet, congrats, good luck in the new company!
20:30:30 <Piet> Yeah, I lucked out because they didn't ask David about me
20:30:41 <david-lyle> #truestory
20:30:53 <tsufiev> Piet, david-lyle  are you going to work in the same room :)?
20:31:08 <mrunge> pffft
20:31:39 <rhagarty> out of the blue question.. is anyone aware of a plug-in that adds actions to a core horizon table WITHOUT the use of an overrides file?
20:31:41 <Piet> The change will allow me to focus more on the community and choose where I focus (like the Swift panel...)
20:31:58 <r1chardj0n3s> \o/ swift panel
20:32:10 <neillc> looking forward to your input :)
20:32:23 <Piet> tsufiev Not sure I want to work at David's house
20:32:33 * david-lyle has to step away for a sec
20:32:39 <tsufiev> Piet, oh, didn't know that
20:33:28 <Piet> tsufiev would like to rent a small studio downtown Fort Collins
20:33:45 <mrunge> rhagarty, not that I'm aware of
20:33:57 <mrunge> rhagarty, I'd think that's impossible
20:34:14 <doug-fish> rhagarty: I don't think we have that capability yet, I expect it to happen for angular based code, something like the workflow extensibility that we've recently introduced
20:34:53 * david-lyle back
20:35:30 <david-lyle> rhagarty: no way to do that now other than override the view/table in your plugin
20:36:17 <rhagarty> thanks... just afraid of my plug-in colliding with another "overrides" plug-in
20:36:46 <doug-fish> I'm afraid that fear is justified
20:36:48 <david-lyle> rhagarty: well not overrides
20:36:57 <doug-fish> we really do need the better solution
20:37:19 <david-lyle> you can define the view and table to inherit from the existing and merely override the column
20:37:32 <david-lyle> and have your panel displayed rather than the default
20:37:38 <doug-fish> problem is if two overrides try to do that
20:38:29 <rhagarty> david-lyle, maybe I can ping you on this later?
20:38:31 <doug-fish> and that's likely to happen on instances and quotas especially
20:38:41 <david-lyle> rhagarty: sure
20:38:59 <david-lyle> look at the admin view for images as an example
20:39:06 <david-lyle> but feel free to ping me
20:39:11 <rhagarty> thanks
20:40:52 <david-lyle> any other topics today?
20:42:03 <rhagarty> GB vs GiB issue?
20:42:26 <david-lyle> ah yes
20:42:56 <rhagarty> https://review.openstack.org/#/c/243880/
20:43:08 * david-lyle was looking for that
20:43:17 <rhagarty> Cinder wants this, other projects don't feel the urgency
20:43:27 <rhagarty> which leads to inconsistencies
20:43:49 <matt-borland> yeah, it's an important distinction
20:43:57 <rhagarty> especially on Nova panels that display block storage data
20:44:57 <david-lyle> rhagarty: the APIs return bytes, no?
20:44:58 <rhagarty> still working through the issues with multiple patches and bugs
20:45:48 <rhagarty> gigabytes or gibibytes?
20:46:08 <matt-borland> rhagarty, I'd thought we don't get bytes for several of those calls
20:46:17 <mrunge> uhm, what is returned, when cinder is disabled?
20:46:18 <matt-borland> which is part of why people were confused
20:46:19 * tsufiev just found out that gibibyte = 2^30 bytes
20:46:31 <tsufiev> a new thing here every day )
20:46:35 <matt-borland> :)
20:46:47 <rhagarty> there are inconsistencies
20:46:53 <r1chardj0n3s> 2^30?? that's odd :-)
20:47:24 <rhagarty> anyway, just wanted to let everyone know we are working on it...
20:47:45 <matt-borland> thanks for getting that together rhagarty
20:47:47 <david-lyle> rhagarty: so the nova api returns values in GB?
20:47:51 <david-lyle> still confused
20:48:01 <mrunge> rhagarty, yes, exactly my question
20:48:08 <david-lyle> why is it not storing the # of bytes
20:48:17 <david-lyle> then the UI formats however we want
20:48:24 <mrunge> the api we're talking to looks like the same, but returns different values?
20:48:30 <rhagarty> still trying to determine, working with them now
20:48:41 <david-lyle> should all be bytes
20:48:58 <rhagarty> cinder had this same issue... api was saying GB but stored GiB
20:49:19 <david-lyle> so the API was returning GiB units?
20:49:25 <rhagarty> yes
20:49:35 <david-lyle> why would it do that?
20:49:37 <tsufiev> omg
20:49:47 <matt-borland> yep
20:49:49 <SergeyLukjanov> sorry folks, I've just seen question from tsufiev about removing sahara code from Horizon, IMO it should be done after all tests will working (and after M-1 released)
20:49:53 <rhagarty> we had a big customer complaim..
20:49:59 <tsufiev> we should tell the holy cross-project API inquisition group
20:50:25 <david-lyle> nobody expects the holy cross-project API inquisition group
20:50:38 <matt-borland> lol
20:50:41 <david-lyle> SergeyLukjanov: agreed
20:50:50 <r1chardj0n3s> heh
20:50:51 <SergeyLukjanov> david-lyle, ack, thx!
20:51:34 <david-lyle> I could see the client formatting bytes to GiB, but why the API
20:51:36 <david-lyle> oh well
20:52:06 <rhagarty> mrunge, answer to question above - LVM default reference returns GiB
20:52:38 <david-lyle> rhagarty: oh so it's just reporting on, starting to make more sense
20:54:15 <mrunge> rhagarty, my concern is more like: we can not influence that, and it would be benificial for us, if we just could have bytes
20:54:21 <mrunge> and format it to our needs
20:54:37 <mrunge> it'd be the same for all other api's
20:55:06 <rhagarty> but Cinder API is very explicit with GiBs, so we should reflect that in our UI
20:55:20 <matt-borland> yeah, the unit of measure is just reported wrong right now
20:55:24 <rhagarty> and storage admins know the difference
20:56:06 <matt-borland> it's not about changing how the data is passed, we know the value, it's just the wrong label.
20:56:25 <rhagarty> correct, sorry that wasn't more clear
20:57:06 <rhagarty> confusion comes with Nova panels - investigating that now
20:57:23 <david-lyle> ok, keep us posted
20:57:45 <rhagarty> will do... sorry for the tangent
20:58:16 <david-lyle> it's good to get people more aware of what you're trying to do and why
20:58:37 <david-lyle> other than you like 'i'
20:58:38 <mrunge> reading through the patch, does it make sense to wait until all values are changed rhagarty ?
20:58:50 <mrunge> you mentioned quotas and defaults?
20:59:29 <rhagarty> mrunge, that could take awhile. Was hoping to get smaller incremental patches so stakeholders could review
20:59:49 <mrunge> do we have a tracker?
21:00:07 <david-lyle> if the data is GiB anyway, what's the harm in correcting it as many places as we can?
21:00:21 <mrunge> merging the patch will have horizon semi fixed and a bit inconsistent?
21:00:35 <david-lyle> mrunge: but it's all wrong now
21:00:39 <r1chardj0n3s> mrunge: but *correct*
21:00:53 <mrunge> now it's consistently wring
21:00:59 <doug-fish> I'm concerned about changing only some parts to use GiB for storage because I think it implies that if we label storage as GB elsewhere that we are correct
21:01:07 * david-lyle not seeing the upside to consistently wrong
21:01:14 <mrunge> wrong, when it's semi fixed, you don't know if it's right or wrong
21:01:15 <rhagarty> david-lyle, maybe we can take this up after meeting?
21:01:26 <david-lyle> sure
21:01:29 <david-lyle> time's up
21:01:41 <david-lyle> Thanks everyone. Have a great week!
21:01:44 <david-lyle> #endmeeting