15:55:28 <RRSAgent> RRSAgent has joined #audio
15:55:28 <RRSAgent> logging to http://www.w3.org/2014/10/02-audio-irc
15:55:30 <trackbot> RRSAgent, make logs world
15:55:32 <trackbot> Zakim, this will be 28346
15:55:32 <Zakim> ok, trackbot; I see RWC_Audio()12:00PM scheduled to start in 5 minutes
15:55:33 <trackbot> Meeting: Audio Working Group Teleconference
15:55:33 <trackbot> Date: 02 October 2014
15:56:19 <olivier> Chair: joe
15:56:23 <olivier> Scribe: olivier
15:56:27 <olivier> ScribeNick: olivier
15:57:01 <olivier> Agenda+ Review of action items https://www.w3.org/2011/audio/track/agenda
15:57:08 <olivier> Agenda+ AudioWorker progress
15:57:16 <olivier> Agenda+ AudioWorker spinoff issues
15:57:29 <olivier> Agenda+ TPAC and Last Call review requests
15:57:30 <kawai> kawai has joined #audio
15:57:37 <olivier> agenda?
15:57:41 <olivier> zakim, bye
15:57:41 <Zakim> Zakim has left #audio
15:57:45 <Zakim> Zakim has joined #audio
15:57:50 <olivier> zakim, this will be audio
15:57:50 <Zakim> ok, olivier; I see RWC_Audio()12:00PM scheduled to start in 3 minutes
15:57:53 <BillHofmann> BillHofmann has joined #audio
15:57:55 <olivier> Agenda+ Review of action items https://www.w3.org/2011/audio/track/agenda
15:57:59 <olivier> Agenda+ AudioWorker progress
15:58:02 <olivier> Agenda+ AudioWorker spinoff issues
15:58:06 <olivier> Agenda+ TPAC and Last Call review requests
15:58:29 <olivier> rrsagent, pointer?
15:58:29 <RRSAgent> See http://www.w3.org/2014/10/02-audio-irc#T15-58-29
15:58:55 <Zakim> RWC_Audio()12:00PM has now started
15:59:02 <Zakim> + +1.408.330.aaaa
15:59:15 <olivier> Previous meeting minutes -> http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0254.html
15:59:41 <joe> joe has joined #audio
16:00:16 <Zakim> +joe
16:00:44 <Zakim> +[IPcaller]
16:01:15 <kawai> IPcaller
16:02:27 <Zakim> + +1.650.253.aabb
16:02:48 <Zakim> +cpn
16:02:55 <olivier> zakim, cpn is me
16:02:55 <Zakim> +olivier; got it
16:03:03 <cwilso> zakim, aabb is me
16:03:03 <Zakim> +cwilso; got it
16:03:10 <BillHofmann> zakim, aaaa is me
16:03:10 <Zakim> +BillHofmann; got it
16:03:15 <olivier> zakim, IPcaller is kawai
16:03:15 <Zakim> +kawai; got it
16:05:23 <olivier> joe: starting the call
16:05:47 <olivier> joe: summarises agenda, asks for input on agenda
16:05:58 <olivier> zakim, take up agendum 2
16:05:58 <Zakim> agendum 2. "AudioWorker progress" taken up [from olivier]
16:06:08 <olivier> joe: outstanding questions raised on the list
16:06:16 <olivier> ... lifetime and callback behaviour
16:06:24 <olivier> ... anything else we need to resolve?
16:06:39 <olivier> cwilso: don't think so
16:06:49 <olivier> joe: walking through those two questions
16:06:51 <joe> http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0262.html
16:07:05 <olivier> http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0262.html -> email about lifetime
16:07:17 <olivier> joe: from a developer point of view, seems like the more urgent of the two issues
16:07:31 <olivier> ... question of when an audioworker node ceases to exist and become eligible for GC
16:07:39 <olivier> ... any views?
16:08:08 <olivier> cwilso: the way this works is fairly well define, with terminate for workers.
16:08:19 <olivier> ... you can GC if there are no event handler and no reference
16:08:24 <olivier> ... otherwise you probably can't
16:08:49 <olivier> ... experience we have with scriptprocessors is that would be called n times per second
16:09:11 <olivier> ... the issue is frustrating because for every issue requiring aggressive GC, there are cases where we'd stop processing when we should not
16:10:13 <olivier> joe: reason I included the cases, which come from the spec, is we are trying to implement, at least in theory, native nodes with worker nodes
16:10:35 <olivier> ... not saying we are trying to aggressively automatically GC
16:10:43 <olivier> ... but terminate() does not feel like the answer
16:11:17 <olivier> cwilso: there are 3 potential cases
16:11:30 <olivier> ... the case where from the main thread you say "I'm done, go away"
16:11:51 <olivier> ... that's what terminate does, there are scenarios where you want that
16:11:53 <rtoyg_> rtoyg_ has joined #audio
16:12:02 <olivier> ... second case is where you want to destroy the node from within itself
16:12:08 <olivier> ... decide "i'm done"
16:12:18 <olivier> ... e.g if you were reimplementing buffersourcenode, done playing
16:12:40 <olivier> ... until it's done playing the node itself can't destroy itself
16:12:56 <olivier> ... that's the onaudioprocess, which will be set to no when done
16:13:31 <olivier> joe: if that's the way to go, we need to add case #5
16:13:51 <olivier> cwilso: think that's what #4 does
16:14:08 <olivier> (the 4 cases are from http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0262.html)
16:14:31 <olivier> joe: think someone in the spec we need to add that onaudioprocess is hook that keeps node alive
16:14:48 <olivier> ... and that you must set the callback to no to kill it
16:15:00 <olivier> ... that'd need to happen to address the case in full
16:15:06 <olivier> cwilso: agrees
16:15:22 <olivier> joe: case #4 has additional question, perhaps can be dispensed with
16:15:35 <olivier> ... asking whether there is anything upstream for me that is still active
16:15:51 <olivier> ... is it necessary to implement e.g. convolver or delaynode?
16:16:12 <olivier> cwilso: don't think it is necessary in API
16:16:24 <olivier> ... the way the node itself manages that is set onaudioprocess to no
16:16:32 <olivier> joe: but it might need to revive itself
16:16:36 <olivier> cwilso: possibly
16:16:50 <olivier> ... could use onconnect/disconnect
16:16:57 <olivier> joe: that would work
16:17:25 <olivier> ... just don't know what priority to set
16:17:34 <olivier> ... the tail-time reference
16:17:44 <olivier> ... don't feel it is a show-stopper
16:18:03 <olivier> cwilso: feel it needs to be figured out before we can consider the solution complete, don't think it is blocking anything
16:18:11 <olivier> joe: would block non-native convolvernode
16:18:54 <olivier> bill(?): kind of ugly
16:19:03 <olivier> cwilso: not married to that design
16:19:15 <olivier> ... think these nodes will be long lived with graph evolving around them
16:20:08 <olivier> joe: maybe this issue would benefit from discussion at TPAC - no quorum on the call
16:20:13 <cwilso> zakim, cwilso has rtoyg
16:20:13 <Zakim> +rtoyg; got it
16:20:32 <olivier> resolution: discussion on tail-time reference postponed to TPAC
16:21:01 <olivier> resolution: document the onaudioprocess = no as way to GC a workernode
16:21:26 <olivier> cwilso: need to add a lifetime section to the node spec
16:21:39 <joe> http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0263.html
16:21:39 <olivier> Topic: Timing and Sequencing
16:21:56 <olivier> http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0263.html -> Refining AudioProcessEvent definition of timing and sequencing
16:22:18 <olivier> joe: a lot of this is not very controversial
16:22:24 <olivier> ... we discussed some of it offline
16:22:41 <olivier> ... found one issue creeping out of this, which is question of consistent buffer sizes
16:23:06 <olivier> cwilso: spec today has block size of 128
16:23:18 <olivier> ... either stick to that or explicitely make block size completely flexible
16:23:25 <olivier> ... big impact on how latency works in the system
16:23:34 <olivier> ... my take on it ATM would be to keep consistency
16:23:55 <olivier> ... alternative would be lot of work, uncertain gain. We'd gain only cycle limitation with 128 sample delay
16:24:07 <olivier> joe: not even sure we get rid of that, might make problem even worse
16:24:37 <olivier> cwilso: note that one stated goal was reduce CPU consumption, can get same effect by batching blocks, already doing that on some systems
16:25:02 <olivier> ... we e.g ask for 4 blocks at a time
16:25:53 <olivier> ... you'd never get a callback for anything different than 128 at a time
16:26:01 <olivier> ... but system may want to batch several at a time
16:26:08 <olivier> ... and go silent for n times as long
16:26:32 <olivier> ... we are guaranteeing monotonically increasing
16:26:38 <olivier> ... but not consistent interval in real time
16:26:54 <olivier> joe: am hearing you would prefer sticking to 128
16:27:07 <olivier> ... there's more stuck to 128 in the spec than I remember
16:27:28 <olivier> joe: moving to second part of http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0263.html
16:27:37 <js0000> js0000 has joined #audio
16:27:54 <olivier> ... listed 3 things, one of which I want to forget/reword
16:28:41 <olivier> cwilso: implementations would balance minimising latency with avoiding buffer overruns
16:28:58 <olivier> ... problem we have is that there are situations we don't want to make that tradeoff
16:29:13 <olivier> ... if optimising for powerconsumption you may want to keep larger buffer around
16:29:18 <olivier> ... thus increasing latency
16:29:37 <olivier> ... topic for f2f as we have a proposal for that?
16:29:44 <olivier> joe: jer had a proposal
16:30:01 <olivier> ... agreeing to push to TPAC, doesn't affect the API so much
16:30:14 <olivier> agenda?
16:30:22 <olivier> zakim, take up agendum 4
16:30:22 <Zakim> agendum 4. "TPAC and Last Call review requests" taken up [from olivier]
16:30:42 <olivier> joe: matt felt it would be good to go through the bugs
16:30:59 <olivier> ... identify whether to fold into v1, last call or not for each issue
16:31:09 <olivier> ... easy to do in a group if we don't stretch it out too long
16:31:17 <olivier> joe: another area is testing
16:31:28 <olivier> don't have a lot of detail, would need to coordinate on the topic
16:31:31 <olivier> q+
16:31:52 <olivier> joe: think we'd like final draft at TPAC
16:31:59 <olivier> ack me
16:32:06 <cwilso> s/at/after
16:32:48 <olivier> olivier: suggest identifying groups to reach out to
16:32:58 <olivier> ... at LC
16:33:10 <olivier> cwilso: won't have a LCWD at tpac
16:33:16 <olivier> ... given number of issues open
16:33:23 <olivier> ... we are shooting for after TPAC
16:33:38 <olivier> joe: was mor eabout informal request
16:35:56 <olivier> olivier: question on how much we want a clear deck of issues before going to LC
16:36:10 <olivier> ... we want to clear all arch issues, but what about smaller issues and feature requests
16:36:21 <olivier> cwilso: think we would want to only have smaller issues
16:36:27 <olivier> ... easier to decide after our review at TPAC
16:36:31 <cwilso> I'd like to note, BTW, that I will be busy elsewhere Tuesday morning at TPAC
16:36:44 <olivier> cwilso: also want to note I will be booked on TUE morning for AB presentation at the AC
16:37:03 <olivier> ... so will be in Audio WG Monday and Tuesday PM
16:37:49 <olivier> Topic: AOB?
16:37:58 <olivier> joe: any other business?
16:38:21 <olivier> cwilso: one thing to mention - just filed an issue
16:38:30 <olivier> ... consistent request to hook in plugins
16:38:33 <olivier> ... VST support etc
16:38:54 <olivier> ... discussion about that and low level destination hooks
16:39:00 <olivier> ... as part of audioworker design
16:39:17 <cwilso> https://github.com/WebAudio/web-audio-api/issues/358
16:39:34 <olivier> joe: good discussion for TPAC
16:39:50 <olivier> cwilso: should probably create another issue for IO and [??]
16:39:58 <olivier> joe: seems reasonable
16:40:19 <kawai> http://www.slideshare.net/yukiotada
16:40:49 <olivier> kawai: want to talk about it at TPAC - we have built http://www.slideshare.net/yukiotada and have seen issues
16:41:01 <olivier> cwilso: you recorded a demo of that for last web music hackathon
16:41:08 <olivier> q+
16:41:19 <olivier> kawai: the demo was not recorded
16:41:24 <olivier> ... but I can show it at TPAC
16:43:29 <olivier> olivier: mention seeing dev tools for web audio, sounds interesting, have requested to have a link sent to the list
16:43:50 <olivier> ... have seen it in nightly
16:43:56 <olivier> cwilso: think it's in regular builds
16:44:15 <olivier> next meeting in 2 weeks
16:44:22 <Zakim> -BillHofmann
16:44:23 <Zakim> -joe
16:44:24 <Zakim> -cwilso
16:44:25 <Zakim> -kawai
16:44:28 <olivier> rrsagent, make minutes
16:44:28 <RRSAgent> I have made the request to generate http://www.w3.org/2014/10/02-audio-minutes.html olivier
16:44:48 <Zakim> -olivier
16:44:49 <Zakim> RWC_Audio()12:00PM has ended
16:44:49 <Zakim> Attendees were +1.408.330.aaaa, joe, +1.650.253.aabb, olivier, BillHofmann, kawai, rtoyg
16:44:57 <cwilso> Jordan Santell's slide deck from the Berlin hackday is at http://jsantell.github.io/web-audio-tools-presentation/#/
16:45:05 <cwilso> (that's on the Firefox web audio dev tools)
16:45:15 <olivier> rrsagent, make minutes
16:45:15 <RRSAgent> I have made the request to generate http://www.w3.org/2014/10/02-audio-minutes.html olivier
18:14:44 <jernoble> jernoble has joined #audio