13:15:18 <abompard> #startmeeting hubs-devel
13:15:18 <zodbot> Meeting started Tue Mar  6 13:15:18 2018 UTC.  The chair is abompard. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:15:18 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
13:15:18 <zodbot> The meeting name has been set to 'hubs-devel'
13:15:28 <abompard> #topic Roll Call
13:15:32 <abompard> .hello2
13:15:34 <zodbot> abompard: abompard 'Aurelien Bompard' <aurelien@bompard.org>
13:15:43 <abompard> #chair ryanlerch shaily[m]
13:15:43 <zodbot> Current chairs: abompard ryanlerch shaily[m]
13:16:05 <shaily[m]> .hello2
13:16:11 <zodbot> shaily[m]: Sorry, but you don't exist
13:16:21 <mizmo> .hello duffy
13:16:22 <zodbot> mizmo: duffy 'Máirín Duffy' <fedora@linuxgrrl.com>
13:16:28 <abompard> that would be without th [m], shaily
13:16:30 <shaily[m]> .hello shaily
13:16:31 <zodbot> shaily[m]: shaily 'None' <shaily15297@yahoo.com>
13:16:47 <abompard> oh there's 4 of us! yay! :)
13:16:50 <abompard> #chair mizmo
13:16:50 <zodbot> Current chairs: abompard mizmo ryanlerch shaily[m]
13:16:56 <shaily[m]> abompard: is the [m] coming from matrix?
13:16:59 <abompard> shaily[m]: yes
13:18:37 <abompard> #topic Status Updates
13:18:41 <abompard> (no action items)
13:18:47 <abompard> who wants to start?
13:19:07 <abompard> maybe shaily?
13:19:31 <shaily[m]> sure
13:20:30 <shaily[m]> i've been looking into the haystack flask port idea, which i think we could integrate in the next release
13:20:58 <shaily[m]> other than that, i was working on the widget search index, which i wanted to discuss in today's meeting
13:21:26 <shaily[m]> what do we want the search UI to look like, for results coming from widget output?
13:22:00 <fm-hubs> pagure.pull-request.comment.added -- shaily commented on PR #540 on fedora-hubs https://pagure.io/fedora-hubs/pull-request/540#comment-47407
13:22:20 <abompard> I don't know exactly, do we have mockup for that ryanlerch & mizmo ?
13:22:47 <ryanlerch> abompard: i am not aware of a mockup for that
13:22:51 <fm-hubs> pagure.pull-request.comment.added -- shaily commented on PR #540 on fedora-hubs https://pagure.io/fedora-hubs/pull-request/540#comment-47408
13:23:10 <ryanlerch> i kinda just winged it on the basic output i did when i threw it in there
13:23:14 * mizmo looks
13:24:37 <mizmo> garrett did do a quick mockup for search results but it uses a different style for the site chrome
13:24:52 <mizmo> shaily: the basic idea - have two tiers of search results
13:25:05 <mizmo> shaily: top tier is objects that match search terms, sorted in columns based on category: groups / people / projects
13:25:25 <mizmo> do we distinguish between team hubs and project hubs yet? if not, then just two categories: hubs and people
13:25:35 <mizmo> the second tier is feed widget results
13:25:42 <mizmo> so this does a search on fedmsg for the term most likely?
13:26:01 <mizmo> but i dont know, implementation wise, what limits / scope is on what we can do in that kind of search
13:27:04 <shaily[m]> i was thinking we could search across all the widgets in the user's bookmarked hubs (starred, memberships, subscriptions)
13:27:28 <shaily[m]> each widget's output
13:27:34 <mizmo> shaily: so what does a single result from a widget look like?
13:27:54 <mizmo> i think it's easier to think about feed widget single result objects because they have a consistent display
13:28:06 <mizmo> but widgets, the display is pretty custom and requires a specific context to make sense
13:28:23 <mizmo> im not saying it's a bad idea, it would be awesome, im just not sure how to display the results you'd get from widgets
13:28:32 <mizmo> can they be formatted as feed widget items?
13:29:12 <shaily[m]> yeah, same
13:29:13 <shaily[m]> i couldn't come up with anything concrete due to the context issue you mentioned
13:29:18 <shaily[m]> sure, we can format the results however we like
13:30:02 <mizmo> shaily: so if there's a way to convert / translate the output of a widget result => feed widget item, that'd be perfect
13:30:52 <shaily[m]> okay, i'll take an example and try to explain
13:31:02 <mizmo> sure :)
13:32:43 * shaily[m] sent a long message: shaily[m]_2018-03-06_13:32:42.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/bRjRiQFRtyRYuvjgRuHgdCRW>
13:33:01 <shaily[m]> basically, the feed widget item idea
13:33:32 <abompard> shaily[m]: could you repeat your last long message? Matrix has cut it when sending it to IRC
13:33:36 <shaily[m]> i'll implement it for one widget and report back, then we can probably discuss how to go forward with other widgets
13:33:46 <mizmo> sounds good to me!
13:34:07 <shaily[m]> oh, sorry
13:34:11 <shaily[m]> for each widget, we'll decide what gets indexed and what gets returned on a successful match
13:34:12 <shaily[m]> say, for the issues widget we have a list of issues, right?
13:34:21 <shaily[m]> one of these is - "fasshim's attempt at preserving memory causes crashes"
13:34:24 <shaily[m]> so if the user searches for "memory", we return the whole string with some contextual info (whose hub this widget is in etc)
13:35:39 <abompard> thanks :)
13:36:42 <shaily[m]> okay, that's all from my side :)
13:37:21 <shaily[m]> this will also make for a good blog post
13:37:33 <abompard> shaily[m]: +1 :-)
13:37:39 <abompard> who's next?
13:38:30 <abompard> OK I can go
13:39:09 <abompard> I've been working to get the staging instance up and running. It's working now, and I'd like to get the prod deployment going as soon as possible
13:39:25 <abompard> however that depends on external factors, such as the security review
13:39:27 <ryanlerch> abompard: any potential roadblocks on that?
13:39:33 <ryanlerch> ah, i type too slow
13:39:35 <abompard> :)
13:39:52 <ryanlerch> i am going to be outish again this week,
13:39:55 * ryanlerch is on PTO
13:39:58 <abompard> oh
13:40:18 <abompard> please take all the rest you need :)
13:40:58 <abompard> Apart from that I've been working on other things. I've also done some research on Javascript tools that we could use in Hubs one day
13:41:02 <abompard> but not before prod :)
13:41:24 <abompard> I'm thinking of Flow, a static type checker, and Prettier, a code formatter
13:41:34 <abompard> Those would be very useful I think
13:42:08 <abompard> When we get to prod, we'll be out of the "relaxed mode" on unit tests, let it be known ;-)
13:42:49 <abompard> Not much more to report from me
13:43:12 <ryanlerch> awesome, thanks abompard!
13:43:19 <abompard> sure
13:43:30 <abompard> wanna do a status update ryanlerch?
13:43:44 <ryanlerch> abompard: nothing much from me this week sadly
13:43:52 <abompard> aside from "PTO on a sunny island under the coconut tree"
13:43:56 <ryanlerch> was all docs fad + travel
13:44:03 <abompard> okay
13:44:24 <ryanlerch> AUS to EUR is quite a distance
13:44:36 <abompard> yep
13:45:09 <ryanlerch> i think i was in stransit for more than half the time i spent there 0_0
13:48:06 <abompard> Alright, mizmo?
13:48:52 <mizmo> i got nothing!
13:49:08 <mizmo> :)
13:49:31 <abompard> ok !
13:50:00 <abompard> #topic Open Floor
13:54:24 <mizmo> are we doing any kind of user testing at this point? (or ready for it?)
13:55:03 <abompard> mizmo: I don't know exactly but it's probably best to wait for the prod deployment non? This way we'll get real data
13:55:10 <abompard> but I don't know exactly what you need
13:55:48 <mizmo> okay! i was just not sure if we wanted people playing with staging at this pt
13:55:53 <mizmo> but yeh witih real data, is much better
13:57:07 <abompard> okay :)
14:00:50 <abompard> #endmeeting