15:00:08 <krotscheck> #startmeeting storyboard
15:00:10 <openstack> Meeting started Mon Sep 29 15:00:08 2014 UTC and is due to finish in 60 minutes.  The chair is krotscheck. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:11 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:13 <openstack> The meeting name has been set to 'storyboard'
15:00:30 <krotscheck> Agenda: https://wiki.openstack.org/wiki/StoryBoard#Agenda
15:00:37 <krotscheck> Who’s here?
15:02:01 * fungi lurks
15:02:16 <ttx> o/
15:02:23 <krotscheck> Oh good, we’ve got some people.
15:02:39 <krotscheck> #topic RabbitMQ failures.
15:02:49 <fungi> that's all solved now?
15:02:52 <krotscheck> Verified with jeblair last week that the application is now reconnecting.
15:02:59 <krotscheck> So that’s all solved.
15:03:00 <fungi> awesome
15:03:07 <krotscheck> #topic Launchpad Migration
15:03:12 <mordred> o/
15:03:30 <krotscheck> This came up in the infra meeting, and a few questions were raised.
15:04:06 <krotscheck> So last week I did a full import of all the projects that need to be migrated, hit all the snags, and applied patches to handle data edge cases.
15:04:11 <krotscheck> (on a local instance)
15:04:30 <krotscheck> The only outstanding issue is preserving id's.
15:04:52 <krotscheck> Which I _think_ landed?
15:05:04 <krotscheck> Wait, no.
15:05:07 <krotscheck> It’s this one: https://review.openstack.org/#/c/123821/
15:05:42 * mordred goes to look
15:06:11 <krotscheck> ttx: In response to your comment, the changes in writer.py are intentional- we are mapping the launchpad ID’s directly to the db index in storyboard.
15:06:20 <krotscheck> I can change the commit message though
15:06:35 <ttx> krotscheck: right, it just felt like the commit message only mentioned the autoincrement, so I was wondering
15:06:42 <mordred> krotscheck: ah - looks good
15:06:56 <krotscheck> Got it. I’ll get that fixed and the inrement message updated.
15:06:58 <mordred> krotscheck: I like setting the auto increment id as a command line argument
15:07:32 <ttx> I can +2 it if you change the commit message and post the new patchset now :)
15:08:02 <krotscheck> ttx: Urm, lemme finish the meeting first :0
15:08:11 <ttx> ack
15:08:21 <krotscheck> (Also, that one’s buried)
15:08:29 <krotscheck> (In a commit history, so blargh)
15:08:43 <krotscheck> #topic server:prod no longer worsk.
15:09:03 <krotscheck> So ttx ran into an issue with testing things last week, in that the command of grunt server:prod no longer proxies properly.
15:09:25 <ttx> krotscheck: I had no idea fixing it would be so... convoluted. But I guess those are valuable changes anyway
15:09:41 <krotscheck> It _looks_ like the reason is that the grunt proxy attaches the ‘on-behalf-of’ header with an http url, and the API server is now forcing the protocol redirect to https, which makes everything break.
15:09:54 <mordred> oh nice
15:10:13 <krotscheck> So the options to fix it were either: A- make storyboard not do that (which, well, why? That’s actually a sane behavior)
15:10:44 <krotscheck> B- Make every engineer install a self-signed certificate and run a local https grunt server (lots of extra steps in the dev setup)
15:10:56 * mordred shudders
15:11:05 <krotscheck> C- Teach storyboard to do CORS, and just have the js-docs-draft build consume the production API.
15:11:27 * mordred feels like he was just reading a patch for C
15:11:29 <krotscheck> And, well, C is something we wanted anyway, so that UI reviewers don’t have to download and reinstall storyboard every time.
15:11:48 <krotscheck> So that’s what I did.
15:12:05 * mordred woots
15:12:11 <krotscheck> The reason it’s urgent is because it’s blocking UI code reviews, and it was unexpected work.
15:13:09 <krotscheck> There are 7 more outstanding patches to make all that work, but right now they only seem to be held up on a merge conflict.
15:13:24 <krotscheck> And it impacts config, projects-config, puppet-storyboard, storyboard, and storyboard-webclient.
15:13:52 <krotscheck> Any questions on that? Should I paste all the links in here or add them to the agenda?
15:14:05 <mordred> krotscheck: is there an order that things need to land to unbock things?
15:14:15 <mordred> krotscheck: also, do they all share a topic I can search for?
15:14:49 <mordred> I get five things with https://review.openstack.org/#/q/status:open+topic:cors,n,z
15:15:07 <krotscheck> They don’t share a topic, sorry :/
15:15:15 <krotscheck> The order should be listed in the code review.
15:15:18 <krotscheck> Sec...
15:15:29 <krotscheck> All CR’s added to the agenda
15:15:47 <mordred> cool
15:16:50 <krotscheck> Any other questions before moving on?
15:17:04 <ttx> nope clear enough
15:17:16 <krotscheck> #topic Discussion: Story Types
15:17:22 <krotscheck> ttx: ANy progress on a spec?
15:17:39 <ttx> krotscheck: no, got buried under other tasks unfortunately
15:18:01 <krotscheck> Cool, let’s discuss next week.
15:18:05 <ttx> but it's still alive on my TODO list, which means it's not *too* buried.
15:18:07 <krotscheck> #topic Discussion - new
15:18:14 <krotscheck> Any new discussion/design topics?
15:18:35 * ttx needs to write more text to exercise his new mechanical keyboard
15:18:42 * mordred has to step away for a sec - will read scrollback/meeting notes when he gets back and review the cors stuff
15:19:24 <krotscheck> kk
15:19:36 <fungi> or we could just talk about ttx's new mechanical keyboard
15:19:53 <krotscheck> #topic MVP 1.1 - Subscription
15:19:53 <ttx> fungi: love it so far
15:20:21 <krotscheck> So, subscription is currently held up on a few things.
15:20:28 <krotscheck> Firstly, this review; https://review.openstack.org/#/c/104004/
15:20:42 <krotscheck> SergeyLukjanov has explicitly asked (in the comments) that NikitaKonovalov review it as well.
15:20:56 <ttx> krotscheck: yeah, that's why I didn't Workflow+1 it
15:21:48 <krotscheck> So hopefully NikitaKonovalov will get around to it :) (I pinged him last week)
15:22:03 <krotscheck> Also, the worker-daemon management landed (it’s neat!
15:22:08 <ttx> krotscheck: we shouldn't hold indefinitely though
15:22:39 <krotscheck> ttx: Well, it’s been a week. I figure if he hasn’t gotten around to it by next monday we’ll put it in.
15:22:48 <ttx> ok, fair enough
15:23:08 <krotscheck> Next steps on worker daemons is the puppet bits and the config bits.
15:23:46 <krotscheck> The first of those is here: https://review.openstack.org/#/c/113937/
15:24:00 <krotscheck> I abandoned the projects-config one because it was originally proposed to config
15:24:40 <krotscheck> Having said that, https://review.openstack.org/#/c/114404/ is the dashboard version of it, and I think it’s aged out now :/
15:25:03 <krotscheck> But once that’s back on the correct history, it should be mergeable without requiring the workers to work.
15:25:11 <krotscheck> It’ll just show nothing for a while :/
15:25:30 <krotscheck> So I guess my question is: Do we land the UI things before subscriptions are actually functional?
15:26:19 <krotscheck> The API already supports subscribing and reading events, so the UI won’t break if they land, it’ll just… well, people will expect things to work and I anticipate a bunch of people asking “Hey, I clicked on subscribe and nothing happened"
15:27:19 <ttx> hmm
15:27:25 * krotscheck can go either way
15:27:26 <ttx> I guess at this point we can afford it
15:27:41 <ttx> maybe not a general rule though
15:28:32 <krotscheck> So, let’s land them once it’s actually working, so that we get into the habit of not exposing users to half-baked features :)
15:28:44 <ttx> balance the inconvenience of not merging it , vs. the inconvenience of people asking wtf
15:29:20 <krotscheck> Fair point.
15:30:00 <ttx> at this point I'm pretty sure not so many people will wtf, but the balance will change fast (I hope)
15:30:14 <krotscheck> Another random question I have for python experts is whether we want to run the workers as an independent daemon, or whether we want to have them be spawned by the long-running wsgi process so that they’re automatically there whenever storyboard is installed.
15:30:46 <krotscheck> Right now we’re doing the first, but I feel that installing will be way easier if we do the second. I just don’t know if it’s feasible (I don’t know enough about wsgi)
15:30:50 <ttx> I don't have a strong opinion on that -- but I suspect others could :)
15:31:05 <krotscheck> Right, so we’ll leave that question open for anyone who comes back to read the backscroll.
15:31:09 <krotscheck> Onward!
15:31:14 <krotscheck> #topic Project Groups
15:31:15 <krotscheck> No progress.
15:31:21 <krotscheck> #topic Tags
15:31:23 <krotscheck> No progress
15:31:26 <krotscheck> #topic Emails
15:31:28 <krotscheck> No Progress
15:31:36 <krotscheck> #topic Open discussion
15:32:00 <ttx> krotscheck: we should plan to have a storyboard session at the design summit again
15:32:15 <krotscheck> ttx: I agree. Also, UX testing with new things.
15:32:16 <ttx> maybe use one of the very few infra "scheduled slots" for it
15:32:42 <fungi> i'm on-board with that suggestion
15:32:55 <krotscheck> I’m doing a lightning talk on monday called “Lightning Walkthrough: StoryBoard” to drum up some interest and recruit UX victims^C^C^C^C^C^Cparticipants
15:33:01 <ttx> krotscheck: the plan for infra is to have 4 scheduled slots and a full day of meetup
15:33:14 <fungi> since it has an impact across all projects, ultimately, and isn't easily tackled via an ml discussion
15:33:29 <fungi> e.g. running demo on projector, q&a, et cetera
15:33:29 <ttx> so the scheduled slots need to be used for stuff that would welcome outside participation/feedback
15:34:16 <fungi> right, its reach also goes beyond just our developer community
15:34:54 <krotscheck> Treating it like a first-class citizen will also make people realize that they can’t keep ignoring it.
15:35:17 <ttx> fungi: infra didn't post an etherpad for planning at https://wiki.openstack.org/wiki/Summit/Planning yet
15:35:24 <ttx> so I didn't add it :)
15:35:41 <fungi> ttx: oops! i'll do that today. thanks for the heads up
15:35:44 <ttx> I wasn't &0% sure you didn't already have one
15:35:47 <ttx> 100%
15:36:02 <fungi> i haven't seen one, so i have to assume we do not
15:36:26 <krotscheck> Alrightey.
15:36:33 <fungi> but we will, shortly, and in tomorrow's meeting i'll prod meeting attendees to start chucking ideas at it
15:37:12 <krotscheck> fungi: Do you want me to put storyboard migration back on the agenda now that we’ve got the issues from last week worked out?
15:38:51 <fungi> krotscheck: well, jeblair's on vacation so we probably don't want to go trying to coordinate a cut-over while he's not around (then again, maybe we should!)
15:39:34 <krotscheck> Well, let’s put it on there. There’s no guarantee that we’ll get around to it anyway.
15:39:43 <fungi> sure
15:39:44 <krotscheck> And we can always leave it on the agenda.
15:39:59 <krotscheck> Alrightey, anything else?
15:40:06 <fungi> honestly i haven't looked at the agenda yet and have no idea how much we're trying to cram in tomorrow
15:40:11 <fungi> so good enough
15:40:31 <krotscheck> Oh. SB Migration is still on there.
15:40:32 <krotscheck> Well then.
15:40:56 <fungi> wfm
15:41:00 <krotscheck> Alright, if there aren’t any more discussion topics out there i’ll close up the meeting.
15:41:08 <krotscheck> Bueller? Bueller?
15:41:49 <krotscheck> Alright, thanks everyone
15:41:51 <fungi> sounds like a resounding "no"
15:41:53 <krotscheck> #endmeeting