16:02:01 <pollo> #startmeeting
16:02:01 <MeetBot> Meeting started Thu Aug  4 16:02:01 2022 UTC.  The chair is pollo. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:02:01 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:02:07 <pollo> #topic Roll Call
16:02:23 <pollo> I'll give y'all 5mins because I need it too :)
16:02:32 <paddatrapper> \o
16:02:35 <wouter> well, I'm here, obviously
16:03:07 <olasd> o/
16:06:03 <sahilister> hello
16:06:56 <anupaannjoseph[m]> o/
16:07:11 <pollo> #topic DC22 Post-Mortem -- pain points
16:07:30 <pollo> anything that went wrong, or we would like to work on for future events?
16:07:43 <wouter> they seemed to be working pretty well from my POV, but then I wasn't on-site
16:07:45 <olasd> "just in time setup" for hired equipment
16:08:01 <olasd> (and spending the whole event chasing gremlins in the audio pathways
16:08:03 <olasd> )
16:08:18 <paddatrapper> Yeah, equipment coming earlier in the week would have helped those audio issues...
16:08:46 <wouter> yeah, that's true. There were a number of apology slides issued due to audio issues
16:08:49 <olasd> noisy audio inputs on the cameras
16:08:57 <paddatrapper> Also we spoke about it during the conf, but I think running audio directly into Vocto instead of via a camera is the way to go
16:09:16 <pollo> I'd be open to experimenting yes
16:09:18 <paddatrapper> LED screens don't really work for presentations
16:09:30 <pollo> let's not do those again :)
16:09:41 <wouter> paddatrapper: the reason we didn't do that in the past was because of A/V sync issues
16:09:46 <paddatrapper> Surprisingly very few opsis issues
16:09:56 <wouter> paddatrapper: I think if we're going to do that, SReview needs better A/V sync UI ;)
16:10:13 <pollo> wouter: apparently, LCA does that and it works
16:10:21 <pollo> as the delay seems to be consistent
16:10:26 <paddatrapper> wouter: yes, but we have desync anyway. And feedback from LCA was that the offset was typically fixed for a given set of hardware
16:10:36 <wouter> right, fair enough
16:10:37 <olasd> still some mysterious a/v desyncs during some days even though we didn't seem to have load issues on the voctomix machines
16:10:53 <wouter> I wasn't saying "don't do it", merely trying to reiterate why we hadn't done it so far
16:10:55 <olasd> hybrid talks needing a lot of babysitting
16:11:04 <pollo> our OBS not being ansibled was a pain point I had
16:11:21 <olasd> ansibled to do what?
16:11:34 <wouter> olasd: do you think we are ready for more hybrid talks in the future, or is that something that requires a lot more work?
16:11:36 <pollo> loopy. We either need more ansible on that role, or better doc
16:11:55 <pollo> ah good pain point: remote talks are bleh
16:12:07 <nattie> (hi, slightly belatedly)
16:12:13 <pollo> it worked, but no one liked them and they were somewhat boring
16:12:41 <pollo> I think we can do hydrid presentation where someone joins a talk, but let's not accept remote talks again
16:12:41 <wouter> okay, fair enough
16:12:51 <paddatrapper> hybrid talks work better when it is a mix of in-person and remote presenters. A single remote presenter to a room of in-person attendees didn't work
16:12:57 <olasd> I think we should keep hybrid events to BoFs/interactive sessions
16:13:00 <olasd> yeah
16:13:01 <paddatrapper> pollo: +1
16:13:14 <pollo> can we agree on that?
16:13:23 <wouter> I think interactivity is required
16:13:30 <paddatrapper> I think so
16:13:30 <wouter> if it's just "play a video", you can ship a URL to a video too
16:13:36 <olasd> pollo: I don't think we need agreeds, just listing stuff
16:13:45 <pollo> ok :)
16:13:46 <wouter> so I *think* remote talk where the talk is done live *could* work. Did we do that?
16:13:58 <nattie> not this time, i think
16:14:02 <nattie> but that is kind of hair-raising
16:14:21 <wouter> so I think perhaps we could consider that in the future, with the caveat that we might fail if network fails us etc
16:14:24 <paddatrapper> we didn't, but the feedback I got from the attendees/presenters was most missed the hallway track talking to the presenter before/after the talk
16:14:33 <pollo> wouter: that opens a can of possible "shit, something went wrong"s
16:14:35 <wouter> but anyway, we shot that down in the past as well, so, meh
16:14:50 <wouter> paddatrapper: oh, right, yes
16:15:04 <wouter> okay, sure, with that feedback, no more remote talks then, fully agreed
16:15:25 <pollo> anything else?
16:15:50 <olasd> I had more but I can't remember :P
16:16:02 <pollo> #topic DC22 -- Status of published videos
16:16:16 <pollo> tumbleweed isn't here, but AFAIU, videos have been published?
16:16:22 <wouter> I think those are done. I saw tumbleweed also merge the upload for peertube
16:16:33 <wouter> I don't know what the status on YouTube uploads is
16:16:34 <olasd> yeah, they are
16:16:46 <olasd> https://www.youtube.com/playlist?list=PLYUtdmpYPTTL6_iJ3kpFtROmnacNg1R2S has 84 videos
16:16:51 <wouter> but essentially that's just "run a script", so
16:16:54 <pollo> #info final videos have been published on our archive, on Youtube and on Peertube
16:16:56 <wouter> even better
16:17:01 <pollo> kudos to people involved
16:17:13 <pollo> #topic DC22 -- Archival of important bits
16:17:14 <wouter> I think our workflow there works well, not much left to say about that
16:17:25 <pollo> I archived the etherpad
16:17:31 <tumblingweed> Sorry, I was waiting for the meeting, and then forgot
16:17:33 <pollo> I saw paddatrapper did grafana
16:17:47 <pollo> we've sent a badger for speakers to upload their slides
16:17:52 <tumblingweed> +1 on the pain points above
16:17:54 <pollo> are we missing something?
16:18:15 <paddatrapper> Yup, that is all merged
16:18:18 <wouter> re: grafana, are we going to ansible that?
16:18:25 <olasd> "raw" footage has been synced to vittoria
16:18:27 <pollo> that's a future goal
16:18:40 <wouter> okay
16:19:00 <wouter> FWIW, SReview has hooks for prometheus data gathering too
16:19:12 <pollo> interesting
16:19:26 <wouter> (FOSDEM uses that as well, so...)
16:19:50 <pollo> #topic DC22 -- Buying new gear?
16:20:02 <wouter> I didn't sync the SReview logs, but I don't think that's crucial
16:20:03 <tumblingweed> Opsis alternatives!
16:20:07 <pollo> I split that one in 2, global gear and gear for the Brazilians
16:20:14 <pollo> let's start with global
16:20:33 <pollo> yes, we should look into something that does 1080@60fps
16:20:44 <paddatrapper> I would like to experiment with replacing the Opsis boards with Elegato capture cards
16:20:55 <pollo> paddatrapper: can I action you?
16:21:01 <wouter> we could use the same hardware that FOSDEM uses, but that's not free hardware
16:21:01 <paddatrapper> Yup
16:21:14 <pollo> #action paddatrapper to look into an opsis replacement, possibly using Elgato gear
16:21:16 <wouter> (it works quite well though)
16:21:21 <pollo> wouter: what is it?
16:21:28 <wouter> I'm trying to remember the name
16:21:47 <wouter> It's a company that makes all kinds of MPEG4 encoders
16:21:51 <wouter> hardware ones
16:21:56 <wouter> ! blackmagic design
16:22:12 <olasd> paddatrapper: have a reference in mind? there's an event in Paris (10 years of guix) coming up in september, may be worth a shot :P
16:22:13 <pollo> I don,t think they have boxes that do capture + HMDI output
16:22:16 <wouter> https://www.blackmagicdesign.com/
16:22:43 <pollo> otherwise, we already have BM gear, it's expensive but it works
16:22:47 <wouter> well, FOSDEM takes their encoder and wraps it in a wooden (!) box that does some other weird shit, but hey
16:23:00 <pollo> anyway, paddatrapper has been actioned
16:23:08 <wouter> that bit is open, I think they have a git repo on github
16:23:11 <pollo> we also talked about XLR capture cards
16:23:21 <pollo> to bypass the cameras
16:23:29 <paddatrapper> olasd: heh, we shall see. I'm not sure I'm going to be able to do another international trip this year
16:23:53 <wouter> sorry, what?
16:23:58 <wouter> bypass the cameras how?
16:24:13 <paddatrapper> We used a Focusrite Scarlette for the remote presentation interface. Worked well and is natively supported on Linux
16:24:14 <wouter> oh, for audio
16:24:15 <wouter> right
16:24:21 <paddatrapper> wouter: audio interfaces
16:24:55 <pollo> I think it would be worth it to buy one of those now and play with it at our next event / sprint
16:25:25 <wouter> +1
16:25:35 <paddatrapper> I think so too. Could just do a 2i2 unless we can think of uses for more inputs/outputs?
16:26:18 <pollo> can the same box be used for our hybrid setup?
16:26:20 <wouter> I personally can't, unless you want to do audio mixing in software
16:26:25 <wouter> (which I think would be a terrible idea)
16:26:34 <paddatrapper> pollo: yes
16:26:47 <pollo> then we probably need one with 3 inputs?
16:26:53 <paddatrapper> We were only using 2 inputs and 1 output for that
16:27:07 <paddatrapper> 1 input and 1 output actually
16:27:22 <paddatrapper> Hmm... I'm not sure we will be able to use the same interface for both at the same time
16:27:39 <pollo> ok, that was my question. We can have 2 boxes, but 1 would be better
16:27:44 <paddatrapper> Otherwise we'll need an HDMI input to the projector from the back
16:28:08 <pollo> then let's get 2 of those?
16:28:25 <paddatrapper> And will have to run Jitsi on the Vocto machine
16:28:28 <paddatrapper> +1
16:28:44 <pollo> olasd: did you have other gear we needed to buy for the EU kit?
16:29:00 <olasd> not that I can think of
16:29:43 <pollo> and would you be ok with the idea of adding up to 4 Focusrite boxes to our gear?
16:29:54 <pollo> I guess they do take space...
16:31:05 <paddatrapper> If we can reduce the size of the HDMI capture boards, they can take that space :P
16:32:12 <olasd> I don't understand your count of 4, but yeah, should not be a problem
16:32:32 <pollo> 1 for each room + 1 for hybrid events
16:32:37 <pollo> #info we are looking to buy 2 Focusrite Scarlette and 1 TBD HDMI capture box soonish
16:32:48 <pollo> ok, now for the Brazillians
16:33:08 <pollo> so they've asked to have a kit for a single room, as they have lots of local events
16:33:15 <pollo> IMO they need something easy to use and simple
16:33:17 <olasd> (for the EU kit we could probably just hook up to the mixing desk via USB)
16:33:33 <olasd> (at least for output)
16:33:43 <paddatrapper> True, makes things easier
16:33:52 <pollo> we took back the opsis they had, since they weren't really using it
16:34:15 <pollo> do we want to meddle in this, or do we let them ask on their own and let the DPL deal with that?
16:34:33 <wouter> I would go for the latter
16:34:57 <wouter> we can guide them as to what works and what doesn't, but it's something they need to understand, so it should be something that they helped build, I think
16:34:59 <paddatrapper> We could recommend gear, but ultimately they need to find things that work for them
16:35:00 <wouter> no?
16:35:03 <paddatrapper> heh snap
16:35:07 <wouter> right :)
16:35:10 <olasd> it makes sense to have a look at what they intend to purchase and provide feedback, but yeah, what you both said
16:36:10 <pollo> #agreed we can provide feedback on an eventual kit for Brazil, but we will let them make a proposal first
16:36:19 <pollo> anything else for this topic?
16:36:50 <pollo> #topic DC22 -- Other
16:37:20 <olasd> it'd be best if the setup was supported within our stack, as it's always good to have more people familiar with it, of course. but it shouldn't be a blocker
16:37:38 <olasd> (better have things done than have things perfect, etc.)
16:38:32 <pollo> we had a DebConf!
16:38:40 <pollo> #topic Potential Autumn Sprint
16:38:59 <pollo> we mentioned it would be nice to have a sprint this Autumn to work on a few things
16:39:13 <highvoltage> I thought the sprint would be in spring
16:39:28 <pollo> ah, I understood Autumn since tumblingweed was in SA
16:39:38 <wouter> I think we should talk months rather than things that are hemisphere-specific
16:39:41 <highvoltage> Autum is... March?
16:39:45 <highvoltage> *Autumn
16:39:46 <pollo> oh sorry
16:39:53 <pollo> my bad, October :P
16:40:03 <pollo> or somewhere around that
16:40:09 <olasd> I found another DC22 pain point (duh): 5 days between arrival and start of show may be too short to have our hardware clear customs :P
16:40:24 <tumbleweed> olasd: :P
16:40:39 <pollo> I know I won't come to SA, but I'd be happy to participate remotely
16:40:45 <paddatrapper> especially if India's customs are tricky
16:40:53 <paddatrapper> An Octoberish sprint works for me
16:41:01 <highvoltage> Indian people have assured me that customs will be tricky
16:41:13 <tumbleweed> olasd: also, we really need the hired gear set up before the night before talks start...
16:41:23 <wouter> October would work for me too, but less so if it's too far away, not sure I can take the time off
16:41:27 <tumbleweed> so we can have a troubleshooting day
16:41:32 <wouter> (I was recently told I'm out of vacation days for a while...)
16:41:39 * tumbleweed can October
16:42:06 <pollo> I'm likely to be on unemployment for the next semester, so dates aren't really a blocker on my side
16:42:43 <highvoltage> fwiw I can host at my place if there's nothing nicer / more convenient / etc (just quite a bit out of the way for everyone)
16:42:44 <pollo> is 4 days enough? I'd envision something like Thursday-Sunday
16:42:53 <olasd> thursday-sunday sounds good
16:43:02 <highvoltage> (and I might be doing some construction too but it might be a minor inconvenience)
16:43:16 <wouter> free hands to help!
16:43:17 <olasd> October... sure. Travel, I doubt it
16:43:42 <pollo> what about I send a poll with a few dates options around October?
16:43:50 <wouter> sounds good
16:44:19 <paddatrapper> Sounds good
16:44:22 <pollo> #agreed pollo to send a poll for an eventual October-ish, Thursday to Sunday videoteam sprint
16:44:36 <pollo> anything else?
16:45:06 <olasd> shall we start a pad for an agenda?
16:45:14 <pollo> good idea
16:45:18 <pollo> or a wiki page?
16:45:24 <pollo> we normally use that
16:45:43 <olasd> either works for me :P
16:46:21 <highvoltage> if hardware is needed for the sprint, it's a good idea to order that asap so that it can be shipped in time and pass customs / post office strikes / etc
16:46:25 <pollo> #action pollo to set up a wiki page for the event
16:46:29 <highvoltage> (or DHL it)
16:46:52 <pollo> highvoltage: I think we're waiting for paddatrapper to find a suitable replacement for the opsis, and then we'll make a request?
16:47:02 <olasd> depending on dates I may be able/willing to travel. We'll see.
16:47:39 <pollo> #topic Any Other Business
16:47:40 <paddatrapper> if the one I'm looking for works, it is a 2-5 day turnaround (Takealot), so don't have to worry too much about that
16:47:53 <paddatrapper> If it is something else, it may be more tricky
16:48:13 <tumbleweed> wouter: I need to give you t-shirts :)
16:48:15 * pollo stole CarlFK[m] cat and it's happy pawing on my desk
16:48:26 <wouter> tumbleweed: ah yes
16:48:38 <wouter> tumbleweed: are you in CPT yet? What's the schedule?
16:48:50 <tumbleweed> I'm here, yes
16:49:06 <wouter> cool! definitely need to hook up then, somewhere soon
16:49:57 <CarlFK[m]> pollo: Jpeg!
16:51:11 <pollo> https://share.riseup.net/#Eq14jJZrkP1jQQ38jsKLZg
16:51:17 <pollo> #topic Next Meeting
16:51:22 <pollo> #topic Next Meeting
16:51:32 <wouter> I don't think we need weekly meetings for the time being anymore?
16:51:44 <pollo> is our Thursday 16:00 UTC schedule still ok? should we look into changing it
16:51:58 <pollo> I agree weekly meetings aren't needed, but we should still do a monthly one IMO
16:52:23 <paddatrapper> +1 to both points
16:52:30 <olasd> if the next meeting is the first thursday of september we need to have planned the october sprint by then
16:52:32 <olasd> probably
16:52:53 <pollo> indeed
16:53:11 <paddatrapper> Maybe the 18th of August (2 weeks)?
16:53:35 <tumblingweed> Works for me
16:53:41 <pollo> wfm
16:54:08 <olasd> sure
16:54:09 <pollo> #agreed Next Meeting will be on Thursday August 18th at 16:00 UTC
16:54:15 <pollo> #endmeeting