13:11:08 #startmeeting hubs-devel 13:11:08 Meeting started Tue Feb 6 13:11:08 2018 UTC. The chair is abompard. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:11:08 Useful Commands: #action #agreed #halp #info #idea #link #topic. 13:11:08 The meeting name has been set to 'hubs-devel' 13:11:19 #topic Roll Call 13:11:25 .hello2 13:11:25 abompard: abompard 'Aurelien Bompard' 13:11:36 shaily, are you around? 13:12:40 .hello dufy 13:12:41 mizmo: Sorry, but you don't exist 13:12:47 grr 13:12:51 .hello duffy 13:12:53 mizmo: duffy 'Máirín Duffy' 13:13:35 #chair duffy 13:13:35 Current chairs: abompard duffy 13:13:56 #chair ryanlerch 13:13:56 Current chairs: abompard duffy ryanlerch 13:14:03 because I know you're here :-) 13:14:25 OK it seems to be the 3 of us 13:14:56 There were no action items in the last meeting 13:15:05 #topic Status Updates 13:15:27 Who wants to start? 13:16:24 OK I'll start. I've been working on reviewing existing Pull Requests and packaging Hubs in a way that would be acceptable to deploy on prod 13:16:30 i have one ticket, it/s a css thing, i haven't yet made progress bc of not being able to set up a devel instance locally yet and deadlines on another project 13:16:33 oops too slow 13:16:42 Oh OK 13:16:48 mizmo: is it the halp widget thing? 13:16:58 abompard: yep :( 13:17:13 * sayan is here 13:17:21 sorry there was a power outage 13:17:24 mizmo: OK, the vagrant setup should work now, whenever you can get back to it 13:17:27 hey sayan! 13:17:29 .hello sayanchowdhury 13:17:30 sayan: sayanchowdhury 'Sayan Chowdhury' 13:17:32 ok great :) 13:17:33 #chair sayan 13:17:33 Current chairs: abompard duffy ryanlerch sayan 13:17:56 Is it all mizmo? 13:19:10 pagure.pull-request.new -- ryanlerch opened pull request #556 on fedora-hubs: show descriptions of widgets in the widget select https://pagure.io/fedora-hubs/pull-request/556 13:19:53 OK, I suppose it is. Good luck with your other project(s)! 13:20:01 it is, sorry! 13:20:05 np :) 13:20:34 So, back to packaging: I now have an rpm for Hubs, I've updated the playbook and installed it on hubs-dev 13:20:43 abompard: yay! 13:20:46 awesome! 13:20:50 abompard++ 13:20:50 sayan: Karma for abompard changed to 2 (for the f27 release cycle): https://badges.fedoraproject.org/tags/cookie/any 13:21:00 I've sent an email to the list and I hope that my method of packaging will be sufficient/approved 13:21:20 related to this too -- this is the RFR tincket in infra 13:21:22 https://pagure.io/fedora-infrastructure/issue/6666 13:21:35 hi, sorry i'm late! 13:21:38 I'm investigating Docker right now when I have no PR to review, but I don't think it'll be ready soon 13:21:50 Hi shaily! 13:21:54 #chair shaily 13:21:54 Current chairs: abompard duffy ryanlerch sayan shaily 13:22:10 abompard: what are the next steps in getting the stage set up? 13:22:40 ryanlerch: I was wondering the same thing. I think we need the team to give us these resources 13:23:07 In the meantime, I think we should focus on testing as much as possible the current state 13:23:41 We've been coding without unit tests for a while now 13:23:45 pingou might be able to help is here too 13:23:46 so I'm not very confident 13:24:11 deployment, not tests :) 13:24:28 pingou is having some distractions around IIRC (ie: kid) 13:24:33 abompard: JS unit tests you mena? 13:25:00 ryanlerch: no, the python unit tests. There's been a lot of new features written without associated tests 13:25:05 or just getting the coverage up for the existing unit tests 13:25:09 the js tests are a whole another story 13:25:14 i'll look into writing python tests for wherever they're missing 13:25:34 thanks shaily! I'm particularily interested it widget's cached functions 13:25:46 okay, so that's first 13:26:01 because when there's an exception there we crash the worker and loose the message 13:26:03 hai pingou! 13:26:25 and there's plenty of unexpected cases to have :) 13:26:32 yes 13:26:52 thanks 13:27:06 about search, can you give me some pointers on how to decide between the different search engine libraries. some parameters that we would particularly prefer 13:27:11 i have whoosh set up right now 13:27:17 widget search too 13:27:23 almost there ^ 13:27:34 shaily: to be honest, I haven't looked at your PR yet 13:27:38 but I will 13:28:02 okay, i tried to replicate the existing hubs search functionality using a search engine library 13:28:10 once that's merged i'll start working on widget search using the same setup 13:28:33 i have a decent idea already btw 13:28:50 cool 13:29:06 are you using a backend abstraction library? 13:29:27 that would allow us to switch from whoosh to something else without too much hassle? 13:29:40 (like haystack) 13:29:58 I think that when we get the staging machine we should do a feature freeze and focus on testing. 13:30:13 abompard: agreed! 13:30:40 OK this is a long status update, who's next? 13:30:46 oh, not yet. i saw that haystack was only available for django. i think there must be something for pure python. i'll look into that 13:30:57 i'll go! 13:30:58 thanks shaily 13:31:13 i have mostly been hacking on widgets 13:31:26 got the PR widget for team hubs done 13:31:33 the badges widget is done 13:31:42 there is a new mailing list widget 13:31:58 another new widget that shows the repos on a team up 13:32:00 hub 13:32:18 is that new one for the MVP? 13:32:45 abompard: which new one? 13:32:47 oh wait it's merged already :) 13:32:50 the repo 13:32:59 there's been so many that I forgot :-) 13:33:04 good job ryanlerch! 13:33:04 sorry 13:33:10 i went a little nuts there 13:33:31 :) 13:33:42 ryanlerch++ 13:34:31 also added the fedocal link to the meetings widget 13:35:50 What's next for you ryanlerch? 13:36:55 just filed a quick PR for this 13:36:57 https://pagure.io/fedora-hubs/pull-request/556 13:37:13 it was something that was kinda annoying me not having 13:37:42 niiiice 13:37:46 great idea! 13:37:53 other than that, i think i will do a bit of an ad-hoc review of the interface, files some bugs 13:37:57 so we can triage 13:37:59 yeah 13:38:09 and get into wrting some tests too 13:38:37 Someone mentionned at Devconf that the "My Stream" link on the left menu should be first and "My Hub" second 13:38:57 because that's where you want to be most of the time 13:39:15 (and that's where you land after logging in) 13:39:29 just a detail, but I'm using this meeting to relay :) 13:39:32 abompard: can you file a quick issue for that? 13:39:36 sure 13:39:39 just a oneline? 13:39:42 oneliner 13:39:47 otherwise i will forget 13:40:41 could you please also check that the mobile view is correct? 13:41:01 abompard: yes, it's okay.... 13:41:07 not great, but okay 13:41:09 cool :) 13:41:28 it will take a lot of playing to get it right 13:41:57 yeah, I wonder if we should put the right column below the main column 13:42:01 anyway 13:42:19 yeah -- fine tuning that is a post 1.0 thing i think 13:42:42 absolutely 13:43:05 thats it from me 13:43:16 thanks! 13:43:33 thanks for reviwing my crazy widgets 13:44:43 haha thanks for writing them ryanlerch :) 13:45:29 pagure.issue.new -- abompard opened a new ticket fedora-hubs#557: "The link to "My Stream" should be before the link to "My Hub"" https://pagure.io/fedora-hubs/issue/557 13:45:53 shaily, wanna do a status report? 13:46:59 i was finalizing the issues widget in the past week and adding widget data to the whoosh index 13:47:26 cool 13:47:29 i couldn't find a backend abstraction library though 13:47:41 i just looked again 13:47:47 is django-haystack using something else under the hood? 13:47:54 so should i start incorporating different backends in different branches 13:47:59 1 second 13:48:22 okay i'll look 13:48:32 but my guess is that its just using the python bindings for each search engine 13:48:52 if not, do you feel like you'll be able to write such an abstraction layer, even a very basic one? 13:49:09 i can definitely give it a shot 13:49:09 (not a requirement, just a question) 13:49:12 cool 13:49:38 i'll use haystack's code for reference 13:49:41 sounds fun 13:49:48 good move :) 13:50:17 a haystack-like library for Flask would help a lot of people I'm sure 13:50:20 not only Hubs 13:50:28 yes :) 13:51:00 but be careful if it's too big a task 13:51:18 okay, i'll spend a couple of days trying to see how haystack works and then report back 13:51:19 it's worth trying but don't overcommit 13:51:24 good 13:51:42 thanks 13:52:12 BTW, I'll say it again: please blog more :-) It doesn't have to be perfect. 13:52:48 but it's been 2 weeks since the last one, so please try to write something this week or the next 13:52:58 okay, i will 13:53:01 thanks :) 13:53:28 you'll make the outreachy folks happy :) 13:54:12 (and other people reading your blog or the outreachy planet, including me) 13:54:29 :) 13:54:38 Anything else to add? 13:54:55 btw would you know if there are any opportunities after outreachy for interns 13:55:28 good question. I don't know, let's talk about this after the meeting 13:55:32 okay 13:55:46 that's all 13:55:55 OK. sayan, do you have anything to report? 13:56:04 if not we have 5 minutes of open floor 13:56:26 #topic Open Floor 13:58:05 OK, then meeting's over! 13:58:07 #endmeeting