16:01:17 <RRSAgent> RRSAgent has joined #audio
16:01:17 <RRSAgent> logging to http://www.w3.org/2016/06/02-audio-irc
16:01:19 <trackbot> RRSAgent, make logs world
16:01:21 <trackbot> Zakim, this will be 28346
16:01:21 <Zakim> ok, trackbot
16:01:22 <trackbot> Meeting: Audio Working Group Teleconference
16:01:22 <trackbot> Date: 02 June 2016
16:01:25 <rtoyg_m> +present rtoyg_m
16:02:40 <cwilso> present+ cwilso
16:02:42 <hongchan> +present hongchan
16:02:50 <hongchan> present+ hongchan
16:02:53 <cwilso> +present cwilso
16:03:03 <padenot> +present padenot
16:03:20 <MikeAssenti> +present MikeAssenti
16:06:25 <kawai> kawai has joined #audio
16:06:40 <rtoyg_m> Pick a scribe victim?
16:06:55 <hongchan> https://github.com/WebAudio/web-audio-api/wiki/AudioWorklet-Examples
16:07:16 <joe> Ray can you scribe? (I'll scribe when you talk)
16:07:26 <rtoyg_m> Yes.
16:07:28 <joe> thanks!
16:07:33 <rtoyg_m> scribenick: rtoyg_m
16:08:01 <rtoyg_m> hongchan: Issues came up in writing examples.
16:10:46 <rtoyg_m> mailing list thread: http://lists.w3.org/Archives/Public/public-audio/2016AprJun/0058.html
16:11:17 <rtoyg_m> joe:  What additional things do we need?
16:11:41 <rtoyg_m> joe: How do audioparams get initialized from the main thread side when instantiating a worklet node.?
16:12:57 <rtoyg_m> joe: Need sequence diagram to show sync and async handling?  How do these work on a worklet?  Not nailed down in examples.
16:13:22 <rtoyg_m> joe: Can you subclass audioworklet node?  Would be useful.
16:13:49 <rtoyg_m> joe: VU meter example shows this.
16:14:28 <rtoyg_m> joe: Would optimize for this use case because worklets (probably) would be in libraries.
16:15:23 <rtoyg_m> joe: How should the packaging issue be handled?
16:15:44 <rtoyg_m> hongchan: Subclassing should not be a issue since we already allow this.
16:16:32 <joe> https://lists.w3.org/Archives/Public/public-audio/2016AprJun/0062.html
16:17:01 <hongchan> https://lists.w3.org/Archives/Public/public-audio/2016AprJun/0073.html
16:17:04 <rtoyg_m> hongchan: Seems overkill to do subclassing to handle this case.
16:17:07 <joe> https://lists.w3.org/Archives/Public/public-audio/2016AprJun/0073.html
16:18:20 <rtoyg_m> joe:  These are two approaches, not exclusive.
16:19:00 <rtoyg_m> hongchan: Doesn't like two approaches having two separate files to define worklets.
16:19:11 <hongchan> https://jsfiddle.net/cjcskv7u/2/
16:19:17 <rtoyg_m> hongchan: Having one unified place make more sense.
16:20:37 <rtoyg_m> hongchan: Proposes new approach in jsfiddle link by defining two classes when registering the worklet node.
16:20:43 <rtoyg_m> joe: Any comments?
16:20:54 <rtoyg_m> ian:  is looking over the ramifications.
16:21:23 <rtoyg_m> joe: Like everything in one class, but makes it impossible to name things in the global scope to instantiate it.
16:22:34 <rtoyg_m> joe: Current approach prevents subclassing which we do want.
16:23:20 <rtoyg_m> joe: With single file, it should import the necessary scripts and do the required things.
16:24:24 <rtoyg_m> joe: Node really needs to live in the main scope.
16:24:46 <rtoyg_m> ian: There are ways to get this to work, but might be too confusing for developers?
16:25:14 <joe> s/this/hongchan's example/
16:26:31 <rtoyg_m> ian: Not something we want to do because it can leak state horribly.
16:26:44 <hongchan> hongchan has joined #audio
16:28:39 <rtoyg_m> joe: With all other nodes, they're ready to use from the main thread (even if audio thread side isn't).  Worklets should behave the same.
16:29:02 <rtoyg_m> joe: Is this possible?  Instantiate processor in main thread?
16:30:02 <rtoyg_m> joe: With that, original design is ok.
16:30:47 <rtoyg_m> joe: To summarize, when you instantiate a worklet node the constructor will instantiate the processor class and won't return until it's done.
16:31:45 <rtoyg_m> joe: Can we pass options arg from node all the way to processor?
16:32:16 <rtoyg_m> joe: Simplifies handling of worklets.
16:33:07 <rtoyg_m> joe: audioworklet would get new options arg to pass to processor.
16:34:07 <rtoyg_m> ian: workletnode arg gets cloned and passed to processor.
16:35:19 <rtoyg_m> joe: Do we need to separate out options between known ones and node specific?
16:35:24 <rtoyg_m> ian: We could.
16:35:32 <rtoyg_m> joe: Which is cleaner?
16:35:48 <rtoyg_m> hongchan: doesn't like two options.
16:36:34 <rtoyg_m> joe: Or one options arg with a data: for other dictionary options for the processor.
16:37:12 <rtoyg_m> ian: Looking to see if anyone else does this....
16:37:36 <rtoyg_m> ian: A single dictionary is fine and the whole thing gets structure-cloned.
16:38:45 <rtoyg_m> joe: Excellent; can we move on?
16:39:12 <rtoyg_m> joe: How do audioparams get declared and initialized.
16:39:28 <rtoyg_m> joe: At the moment declaration is on the processor, but in a later example it's moved to the node.
16:40:03 <rtoyg_m> hongchan: audioparam belongs to node, not processor.  Haven't yet decided.
16:40:36 <rtoyg_m> joe: Another thing:  What if param descriptor array were part of options parameter.
16:41:13 <rtoyg_m> ian: Argues that the should be tightly coupled to processor.
16:41:44 <rtoyg_m> joe:  Agrees. In favor of leaving audioparam on processor.
16:43:58 <rtoyg_m> hongchan: param defined in node, but only processor has access to param contents.
16:45:04 <rtoyg_m> joe: Both sides need to know what params are.
16:45:25 <rtoyg_m> joe: Doesn't matter where they're defined, but the more constrained place is the processor.
16:46:52 <rtoyg_m> hongchan: How is subclassing now possible?
16:47:57 <rtoyg_m> joe: Maybe we should discuss initializing audioparams.
16:48:16 <rtoyg_m> joe: Where does intialization actually happen? Appears magical.
16:52:23 <rtoyg_m> joe: Having super() do the initialization of the audioparams so everything is set up and to the processor.
16:53:11 <rtoyg_m> ian: Doesn't need to be (caching, etc), but fundamentally looks synchronous.  Which is good.
16:53:28 <rtoyg_m> joe: We touched everything and feels like we have a path forward.
16:53:57 <rtoyg_m> ian: processor() is the onlly place with access to audioparams?
16:53:59 <rtoyg_m> joe: Yes.
16:54:10 <rtoyg_m> ian: parameter descriptors should be static.
16:54:32 <rtoyg_m> hongchan: If so, can't access from processor.
16:55:01 <joe> +1 on that
16:55:36 <rtoyg_m> ian: Should work.  Then worklet node magically knows the audioparams, synchronously.
16:55:53 <rtoyg_m> padenot: Generally agrees with the approach.
16:56:11 <rtoyg_m> joe: Should rework the examples using this new approach.
16:56:40 <rtoyg_m> honghan: Needs a lot of work to fix up everything for the spec.
16:56:49 <rtoyg_m> joe: can help, but lets work on the examples.
16:57:05 <rtoyg_m> joe: Let's keep pushing ahead.
16:57:23 <rtoyg_m> joe: Any other business?
16:58:15 <rtoyg_m> joe: Adjourn.  Really good call.
16:58:42 <joe> rrsagent, make minutes
16:58:42 <RRSAgent> I have made the request to generate http://www.w3.org/2016/06/02-audio-minutes.html joe
16:59:27 <hongchan> hongchan has joined #audio
17:08:44 <hongchan> hongchan has joined #audio
19:33:05 <joe> action: Joe and Hongchan to update examples and IDL to reflect today's discussion
19:33:05 <trackbot> Created ACTION-132 - And hongchan to update examples and idl to reflect today's discussion [on Joe Berkovitz - due 2016-06-09].
19:33:14 <joe> rrsagent, make minutes
19:33:14 <RRSAgent> I have made the request to generate http://www.w3.org/2016/06/02-audio-minutes.html joe