18:04:51 #startmeeting 18:04:51 Meeting started Wed Jul 1 18:04:51 2015 UTC. The chair is franklin. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:04:51 Useful Commands: #action #agreed #help #info #idea #link #topic. 18:04:59 #chair h01ger edrz wouter tumbleweed 18:04:59 Current chairs: edrz franklin h01ger tumbleweed wouter 18:05:04 #topic Introduction round 18:05:10 Who are you? Experience? Which software stack? DebConf "Privileges"? Your plans? 18:06:02 * tvaz ~back 18:06:10 tassia is coming 18:06:13 have been doing fosdem video for a few years, was maintainer of dvswitch until I threw in the towel 18:06:20 (but I think you knew that ;-) 18:06:31 * h01ger sorta waves but also probably will go afk in 3min 18:06:52 Stefano Rivera. DD for several years, pythonista and opsy-person, Video Team since DC11. Hosting DC16. 18:07:03 tvaz is tiago, old ex team member 18:07:21 i'm sorry, etootired… 18:07:22 cate, only as video-team volunteers [dc10,dc12,dc13] [camera, director], access to some debconf machines. 18:07:24 Hello. 18:07:48 * urbec mostly dvswitch and penta style review 18:08:01 * franklin has been involved in Videoteam since DC8 I thing. initially to handle the camera/DVswitch, then helped for setup. I have commit right on videoteam repo, and I'm alioth admin. I am not DD, I don't know much about audio, the software stack. 18:08:04 I am planning to help for setup (arriving for DebCamp). then for the actual conference. 18:08:26 I arrive for debcamp too 18:08:26 * tassia old team member, I'm here to help the setup of the team, probably won't be available during the conference 18:08:35 haven't decided yet whether I'll help. Will depend on needs 18:08:38 I will be there for debcamp, too 18:09:07 * urbec might help at setup 18:09:15 I will participate as volunteer in VideoTeam and also since DebCamp I can help for setup. 18:09:18 I'll give some help as well for dc16 video 18:09:25 \o/ 18:09:52 dc15 i mean, which doens't exclude dc15 :) 18:09:56 16 18:09:58 oh my 18:10:03 i got it 18:10:06 you 18:10:07 shit 18:10:12 maxy, RichiH CarlFK , what about you? 18:11:01 h01ger, bevevolant dictator this year ? 18:11:18 I think he's already announced he wasn't going to do that 18:11:28 mostly lurking. have been doing videos with dvswitch about 5 years. I think there are about 2000 produced that I was involved with 18:11:55 #topic Who is in touch with orga-team? 18:11:59 What has been discussed / should be discussed? 18:12:51 I'm in orga team, and I think there should be one or two from "content" team, for coordination 18:12:56 RichiH and I (shadow) lead the infra team. But we never really got the team rolling 18:13:06 maxy: ^^ 18:13:20 and jathan ^^ 18:13:50 we could start talking about the gear 18:14:00 #info there should be one or two from "content" team, for coordination 18:14:02 Yes 18:14:05 where it is and what is the plan to bring it to heidelberg 18:14:14 #topic Hardware 18:14:21 Two subjects to address: What we have/where is it? then What we need? 18:14:24 What we have/where is it? (https://wiki.debconf.org/wiki/Videoteam/HardwareInventory ?) 18:14:40 most of it is in Paris, I presume? 18:14:45 I've never really known, but I presume too 18:14:55 And I think h01ger usually brings some of his own things? 18:14:59 I can pick up one or two items from FOSDEM, and put them in my car 18:15:05 and some old list that shoudl be also updated https://wiki.debian.org/Teams/DSA/non-DSA-HW 18:15:12 mostly things like twinpacts 18:15:22 (FOSDEM still has them, doesn't use them anymore) 18:15:35 #info Hardware seems to be in paris 18:15:50 #info wouter can bring some FOSDEM stuffs 18:15:51 to really discuss that, we need to talk about the stack we're going to use 18:16:10 what are the options? 18:16:11 I assume dvswitch + twinpacts, because it's too late to switch to something else 18:16:12 I think for this year, it's way too late to switch to a different stack 18:16:24 but we really need to play with some new things at debconf 18:16:28 or we'll never switch to anything else 18:16:29 yeah 18:16:47 OR we need to have an in-person sprint 18:16:50 I can talk to the FOSDEM team and maybe bring one of their BlackMagic thingies 18:16:57 which we talked about but never made a decision on... 18:17:16 FOSDEM bought 25 or so of those, but they're planning all kinds of weird things involving banana pies etc 18:17:23 is there still time for such a in person meeting? 18:17:32 I don't think so 18:17:32 tassia: not before this dc, I think 18:17:47 we can play at debconf with gear, though 18:18:10 what was the last stable dvswitch we used? Our repo on wheezy? 18:18:18 Shall we discuss the software stack now, and come back to the hardware later ? 18:18:32 dc14 used CarlFK's stuff that was a mix of ubuntus and custom-built stuff 18:18:34 I think the two are related :) 18:18:54 #topic Software stack and related hardware 18:19:38 * nattie waves belatedly 18:19:50 Intro for newbies (if anyone read the logs): The software stack typically address: live production, managing volunteer shifts, review, encoding/publishing. We also pull data from DebConf schedule. 18:20:08 mithro will be at DC15 with the new fpga based hdmi capture board 18:20:17 Is there some part of the stack which are already decided ? 18:20:32 CarlFK: are you actually using that yet? You had a demo one at dc14 that didn't seem usable, yet. 18:20:47 * MadameZou back 18:20:57 #save 18:21:10 I'm working on the managing volunteers, porting the fosdem-volunteer 18:21:23 I think all this talk about future gear is not very useful for *this* dc15 (and hence this meeting) 18:21:58 wouter++ 18:22:02 I am not. currently it fails about 20% of the time to send a good EDID to the presenters laptop. result is the presenters laptop doesn't know what sort of device is plugged into it's hdmi port, and so sending video to it fails. 18:22:16 wouter ++ 18:22:35 fair point. But if we're going to actually get a feel for new equipment, we may want a room using it 18:22:42 could we use new stack in one room, as preparation for next year? 18:22:46 tumbleweed: no, I don't think that's a good idea 18:22:53 wouter: not a primary room - a bof room 18:22:54 we should bring the gear, and play with it in the evening 18:23:06 use our tried-and-tested stack in the rooms where we actually want to record things 18:23:11 absolutely 18:23:13 (also, just ftr my presentation: team member from a couple of debconfs and miniconfs, available for grunt work :) [camera, director, audio], some reviewing experience, subtitling experience, will be there for camp and conf] 18:23:52 so a debcamp sprint for videoteam or something 18:23:58 tumbleweed: I think it's fine if we break down a room for an evening, hook up the cameras to the new playthings, test, and then hook it back up to the old gear 18:24:03 wouter: part of the reason we're so disorganised this year is that we all know we have to move from dvswitch, but we can't yet 18:24:10 so that things work again 18:24:29 Yes. We need to do that long-term. Short term, dvswitch Just Works(tm) 18:24:31 mostly 18:24:32 that is fine, but it wouldn't give me the confidence to rely on it in the next year 18:24:49 I don't want to rely on it next year either 18:24:52 tumbleweed, we could have a sprint just after dc15? 18:25:11 that's normally a fairly high burnout time :) 18:25:14 but yes 18:25:17 I'm just afraid that we don't have thing ready for dc15 if we start testing new solutions before 18:25:45 tassia: I think we're all agreed that we don't want to be relying on anything new for dc15 - we just aren't ready for that 18:25:50 yeah 18:26:01 I think we should bring the gear with the expectation that maybe it will be for nothing 18:26:08 well, we'll probably be using a new encoding pipleline and volunteer system 18:26:13 but those are non-critical 18:26:15 I don't think much about trying new things needs to be planned in a meeting like this. at most a list of things that will likely show up so people can investigate if they are bored. 18:26:19 but if we do find the time to play with those things and maybe have a video team bof where we can play with it, why not? 18:26:24 #agreed we don't want to be relying on anything new for dc15 - we just aren't ready for that 18:26:37 CarlFK: exactly 18:26:47 what('s important is how to get the gear in Heidelberg 18:26:51 s/(// 18:27:01 tumbleweed, i think she meant a risk of not having the regular setup ready for dc15 once we spend time trying new stuff during debcamp 18:27:02 who is in Paris and could bring stuff? 18:27:03 if we have the option to get a couple of extra HDMI cameras + a mixer etc. for playing with, we should 18:27:08 that's the only reason I'm harping on about this 18:27:25 sure, when the minimum setup is ready, we can open a playground room 18:27:35 but we have to plan ahead, if we're going to have extra hardware 18:27:50 Sorry for not talk too much. I need to follow you first in the context and being getting involved with the arising stuff. 18:28:15 But I will be here and in mailing lists 18:28:34 tumbleweed, so bring the extra hardware 18:28:57 would we need to buy new stuff for those tests? 18:29:00 wouter: yeah, try get that capture thing 18:29:08 tassia: I'd hope not 18:29:22 back to: < wouter> who is in Paris and could bring stuff? 18:29:40 franklin? 18:29:41 someone in -team wrote about it. We need to ping them again 18:29:42 $shiping company inc. 18:29:56 tumbleweed, I'm in Paris, but i planned to go by plane :-( 18:30:00 is there already any 'new video gear 2.0' wiki page around? 18:30:03 CarlFK: that's of course always possible, but it's preferable if we don't need to do that 18:30:10 tvaz: nope 18:30:27 People from paris come with many cars, so I don't think it is a problem. We need only to coordinate, maybe a mail in -team? 18:30:35 let's organize one then, and so back to real video situation topic for dc15 :) 18:30:38 I would plan and budget for it. it is easy to change at the last minute if you find someone to take advantage of :p 18:31:20 the budget is already quite far advance. We've been tardy in telling them our needs :( 18:31:23 *advanced 18:31:39 yes, okay so 18:32:07 I suggest we appoint one person to deal with that: talk to "people in paris", see who's coming and could bring stuff, see what's missing, organize shipping 18:32:25 oterhwise we're going to be discussing this on this channel in circles with no appreciable progress 18:32:35 I see we have €5k budgeted for us, but I think we are supposed ot use less 18:32:38 any takers? ;-) 18:32:55 tumbleweed: yes, that's absolutely the case 18:33:03 I can do it, if nobody of core team would make it 18:33:20 we have other things in paris (hopefully) for front desk 18:33:38 cate: thanks! 18:33:39 thanks cate 18:33:52 #action cate will find volunteers to bring the hardware from Paris 18:34:07 should we try to audit what we have in paris? 18:34:10 My flight will take me to Paris first in Charles de Gaulle airport during the first days of August, before August 5, so I can help if you want with something. 18:34:23 wouter, I guess you can bring the stuff from FOSDEM?:-) 18:34:29 I've never been involved in pre-dc shipping plans. I don't know who knows what :( 18:34:42 franklin: I'll talk to them, yes. It's not at my place anymore (used to be), but it shouldn't be an issue 18:34:46 https://wiki.debconf.org/wiki/VideoTeam/NewGear 18:34:51 jathan: with flights is a lot more complex 18:34:54 I can coordinate with someone and help to carry things to Heidelberg. 18:34:57 especially bringing the twinpacts should be a non-issue, since fosdem is done with those 18:34:57 do we know how many rooms we can handle with that hardware? 18:35:03 franklin, please tag info https://wiki.debconf.org/wiki/VideoTeam/NewGear 18:35:12 #info https://wiki.debconf.org/wiki/VideoTeam/NewGear 18:35:31 realted https://wiki.debconf.org/wiki/Videoteam/Hardware and https://wiki.debconf.org/wiki/Videoteam/HardwareInventory 18:35:51 #info also https://wiki.debconf.org/wiki/Videoteam/Hardware and https://wiki.debconf.org/wiki/Videoteam/HardwareInventory 18:35:58 Yes, I only will arrive by plane from Mexico to France and from here maybe I will take a train or go by car withsomeone if it is possible. 18:36:11 tumbleweed, do you know how many rooms we can handle with that hardware? 18:36:22 tassia: usually 2 18:36:30 2 big cameras, 2 small cameras, 2 twinpacts 18:36:32 Do we need more hardware? (typically what h01ger used to bring) 18:36:39 and IIRC we have 1 mixer, and 1 h01ger mixer, usually 18:36:47 I think people are expecting more 18:36:56 fosdem used to rent mixers for €15/weekend 18:37:01 I don't think that's a problem 18:37:08 should we ask what material is available on the venue? Do you have something specific to ask? 18:37:14 I mean mainly audio 18:37:21 we often hire some audio equipment, yes 18:37:30 and we often use venue mixers, with varying success 18:37:30 cate: good point 18:38:18 h01ger, Will you be able to bring some of your stuff? 18:38:25 who is close to the venue? IIRC RichiH might be? 18:38:44 #help we should ask what material is available on the venue 18:38:53 there's one DD in Heidelberg 18:39:00 I slept at his place last year :) 18:39:07 mario... something 18:39:10 * wouter and names... bad combination 18:39:26 DLange: could you take care about this? ^^ 18:40:00 #nick DLange 18:40:08 I'd imagine the venue can also suggest a local equipment hire place 18:40:47 * nattie can look at this, too, if it helps 18:41:05 i see a camera rental place in Mannheim, for one 18:41:33 hiring DV cameras is getting 'hard' 18:41:36 nattie: let's talk also with local people 18:41:57 #action DLange check the Audio stuffs available for the conference 18:42:09 #action franklin make sure DLange accept the task! 18:42:36 franklin: better not to assign tasks to someone who isn't active :) 18:42:37 tassia: who must we talk to about room coverage? 18:42:58 I mean, I don't think we can promise more than 2, unless we find cameras for hire 18:43:05 and laptops, with firewire 18:43:08 not easy :/ 18:43:53 Let's plan for 2, then if we get materials... 18:43:58 I think maxy sent a message some days ago 18:44:01 wouter, I agree. If DLange can't do/don't want to handle that, I'll come back to the list 18:44:46 tumbleweed, but it was a request from the content team 18:44:55 Could we discuss the extra room later, and make sure we have all the hardware for those two rooms ? 18:45:00 franklin++ 18:45:09 yes 18:45:16 BTW there is also an other logistic problem. 18:45:50 plenary room is small, so local team want that video team stream plenaries also in other rooms 18:46:13 but this we can discuss later. And after getting the blueprints 18:46:32 cate: that should be easy 18:46:46 we already stream, that only needs a laptop or something hooked up to a projector 18:47:20 yeah. It just requires higher reliability than we often get in the first few days 18:47:27 really they want also video coverage from questions from such rooms, but I don't think it is easy feasible 18:47:32 #idea local team want that video team stream plenaries also in other rooms ( plenary room is small) 18:47:57 cate: that's not impossible, but I wouldn't promise it 18:48:14 if we have an extra camera, it should be feasible 18:48:17 people in other room could use IRC (not really nice, I know) 18:48:17 But I think the priorities now are to setup the video team/stack/hardware 18:48:19 and an extra laptop 18:48:20 yeah 18:48:23 audio would be a bit painful 18:48:36 cate++ 18:48:36 tumbleweed: mm. Audio isn't my strong suit. Maybe. 18:48:47 cate: dvswitch works fine on wheezy, but is broken on jessie 18:48:53 so we should make sure the laptops run wheezy 18:49:00 since they're ooolld, that shouldn't be a problem 18:49:04 :P 18:49:13 I can cover that 18:49:14 I can maybe also bring a few other laptops, but no promises there 18:49:21 #info dvswitch works fine on wheezy, but is broken on jessie 18:49:24 yes, we should always bring a few spares 18:49:55 I can help checking the Debian version in that laps and install Wheezy in the missing ones 18:50:06 With others or only me 18:50:06 anybody willing to on the FAI stuff for dc15? 18:50:13 jathan: we have automated tools for bootstrapping the machines 18:50:21 Ah ok 18:50:22 FAI 18:50:27 CarlFK: do you remember well enough how h01ger's FAI stuff worked for FOSDEM '14? 18:50:32 tvaz, I can help 18:50:33 I do 18:50:40 I did a lot of work on FAI at dc13 18:50:43 tumbleweed: Thanks. 18:50:52 wouter: I never was able to get h01ger's FAI stuff working at home 18:51:05 tumbleweed, cool then, i did it once but i'm not too familiar 18:51:06 CarlFK: okay, so it falls to tumbleweed then -- I have pretty much the same issue as you :) 18:51:13 franklin, cool aussi 18:51:28 what we never FAIed is the h.264 streams 18:51:44 we did h.264 streams? 18:51:53 for the last 2 DCs, yes. 18:52:03 interesting. Never noticed. 18:52:14 I do have a pretty hardened pxe/preseed install process, complete with an install.sh to boot strap the pxe server https://github.com/CarlFK/veyepar/blob/master/setup/nodes/pxe/install.sh 18:52:15 We *could* also switch to flumotion. 18:52:22 I used that a few times for FOSDEM, and it worked well 18:52:33 this may have had something to do with the fact that flumotion sponsored FOSDEM though 18:52:43 and they sent someone over to keep an eye on things 18:52:52 so when things broke, it never was brought to my attention 18:52:54 I used flmotion at dc14 18:53:38 I can update FAI to install the laptop with dvswith for DC15 18:54:28 I need help for : H264 and a DD wither privileges for UD 18:54:31 I'll commit to doing some work on FAI before DC15, too. And try to get the h.264 stuff set up in advance, this time (rather than during DC) 18:55:15 great 18:56:20 #action tumbleweed and franklin to hack on FAI 18:56:23 what else do we need to talk about? 18:56:39 next topic (?) 18:57:00 Can wee agree on the various software we use? 18:57:19 please 18:57:22 live production? 18:57:35 managing volunteer shifts? 18:57:39 review? 18:57:48 encoding/publishing? 18:57:59 live production is going to be dvswitch, we already agreed on that 18:58:01 live production: dvswitch 18:58:13 #agreed live production: dvswitch 18:58:16 volunteer shifts, I think cate said he was going to look at the fosdem volunteer thingy 18:58:19 streaming: icecast, cubemap 18:58:24 cate: ^^ did I get that right? 18:58:27 how are you managing the dvswitch clients? 18:58:32 yes. 18:58:41 CarlFK: eh? 18:58:49 cate: do you need help on that? 18:58:52 tumbleweed: what starts dvsouce-firewire ? 18:58:53 #agreed cate will look at fosdem volunteer app 18:59:03 CarlFK: the x session on the dvswitch laptop 18:59:08 #agreed volunteer shifts: use Cate's new tool 18:59:10 CarlFK: our FAI thing does all of that from xsession. I'm tempted to use your python thing, though 18:59:30 franklin: it is not my tools. I just deploying for testing into our debconf tools 18:59:58 franklin: that's what I said :) 19:00:01 the tools is coded by Nighty of FOSDEM 19:00:06 https://github.com/CarlFK/dvsmon <- python thing 19:00:08 cate: Juan Rial, fwiw 19:00:21 and MarkVDB, too 19:00:30 details 19:00:39 review: we can't reuse the penta thing 19:00:48 wouter, Yes I'm talking to MeetBot 19:00:56 franklin: my point was, so did I :) 19:01:03 you made me chair, so that gets recorded 19:01:07 ok 19:01:23 cate, please add that on https://wiki.debconf.org/wiki/VideoTeam/NewGear 19:01:37 #chair tvaz 19:01:37 Current chairs: edrz franklin h01ger tumbleweed tvaz wouter 19:01:42 tvaz: I think that's orgogonal 19:01:49 I understand CarlFK's review thing was used for dc14 19:01:52 how well did that go? 19:02:05 it was used for FOSDEM14 as well, not entirely without issues 19:02:15 it relied on him putting in a lot of post-conference work to get the videos done 19:02:24 (I guess I'm mostly asking whether it's a good idea to do that again) 19:02:29 historically, we've always tried to do it before the end of teardown 19:02:39 yes, that was the main problem with FOSDEM, too 19:02:41 um.. i didn't do much at all. nattie did ;) 19:02:46 :) 19:02:47 well, yeah 19:02:56 the main problem was that "loads of work" needed to be done post-conf 19:03:11 which was one of the reasons why FOSDEM14 was a big fail, video-wise 19:03:29 reviewing on site is a plus for me 19:03:42 quality assurance 19:03:50 CarlFK: are you confident that you'll be able to set up things so that people can review (in the evening) what has been recorded that same day? 19:03:53 i think reviewing on site works if you can get people to commit to it 19:03:55 if not, I think veyepar is a no-go for me 19:04:04 which if you do it straight away is pretty painless 19:04:49 or, you know, as close to straight away as possible 19:04:50 the thing is that if we don't manage that. Trying to get anyone to do anything in the month or two after DC is hard. They're all exhausted 19:04:53 except apparently nattie :) 19:04:58 wouter: yes, given the same requirements of any review system: mainly the reviewers need some sort of access to the files 19:05:09 tumbleweed: oh, i was exhausted 19:05:35 CarlFK: have you ever seen debconf's penta video thing? 19:05:45 basically the way there is that it tries to do as little as possible 19:05:54 #save 19:06:09 you NFS-mount things, claim files on a webinterface, and then (on the same webinterface) categorize those files 19:06:20 as "junk", "talk X", "badly recorded talk X", etc 19:06:29 wouter: a little at dc10 and it looked like more work that veyepar 19:06:38 veyepar is even simpler to use, in that you review from a