- From: Raymond Toy <rtoy@google.com>
- Date: Thu, 29 Oct 2020 10:29:42 -0700
- To: "public-audio@w3.org Group" <public-audio@w3.org>, public-audio-comgp@w3.org
- Message-ID: <CAE3TgXHPH7PUOag7w4dN16GZe-82WCC=ib4ULfRnDimywHR0FQ@mail.gmail.com>
Attendees:
Christoph Guttandin, Paul Adenot, Raymond Toy, Jeff Switzer
Minutes:
- V1
- Audio worklet error handling
<https://github.com/WebAudio/web-audio-api/issues/2267>
- Paul: I think what we do makes sense since we're on a different
thread and want to report the error to the main thread.
- Raymond: The WHATWG PR <https://github.com/whatwg/html/pull/6056>
says "'report an error' to bail out non-EventTarget globals like
WorkletGlobalScope." Not sure what "bails out" means here.
- Paul: Yes, that's what we want.
- [Paul updates and closes issue]
- AudioWorkletProcessors cannot process music
<https://github.com/WebAudio/web-audio-api/issues/2268>
- Raymond: Isn't a small ring buffer what they want to use here? Like
what many worklet demos use?
- Paul: Yeah, or transfer the buffer.
- [Paul updates and closes issue]
- Exclusive access to audio HW
<https://github.com/WebAudio/web-audio-api/issues/72>
- Christoph: It's about the interrupted state that Safari introduced
- Paul: The intent is to notify devs it's been suspended in a
different way than calling suspend().
- Christoph: What's the state of Chrome and Firefox on Android
when a phone call comes in?
- Paul: Firefox suspends the audio.
- Christoph: So both Chrome and Firefox do this?
- Raymond: I don't actually know what happens. I'd have to test
this.
- Paul: Opinions?
- Raymond: Seems useful to let devs know what's happening when the
context is stopped.
- Raymond: Do we want to do this for V1? Lots of detail that
needs to be added
- Paul: There's an additional factor that this has been shipped
(on Safari) for sometime. Seems simple enough and it's not a hack. It
belongs to the state.
- [Raymond moves issue to V1 project, In WG Discussion]
- [Paul updates issue]
- Status of outstanding issues that we should fix for v1.
- Paul: I've found cycles now to work on this. Working on issue 2008
(Integration of event loop and rendering loop)
<https://github.com/WebAudio/web-audio-api/issues/2008> now.
Implementations seem to be compatible and will start writing tests.
- Raymond: I need some help on issue 2248 (ABSN.start sending a
control message on suspended context)
<https://github.com/WebAudio/web-audio-api/issues/2248>; I was
confused on what we're actually trying to do here.
- Paul: I can take a look at that, and issue 2251 (Initial
transition to running must consider [[suspended by user]])
<https://github.com/WebAudio/web-audio-api/issues/2251> which is
closely related.
- Raymond: Still working on frozen arrays
<https://github.com/WebAudio/web-audio-api/issues/1933>; just need
to describe how freeze works for the parameter object.
- Paul: Yeah, lots of ECMAScript text needs to be translated for
the spec.
- Open PRs
- ABSN extrapolates last output
<https://github.com/WebAudio/web-audio-api/pull/2256>
- Raymond: I need to review Karl's comments
- V2
- PannerNode HRTF
<https://github.com/WebAudio/web-audio-api-v2/issues/102>
- Raymond: Just need to prioritize
- Paul: We can have an ontypechange event.
- Raymond: Or be compatible with setBuffer for ConvolverNode.
- Paul: Yes, I was going to say that too. We should be
consistent. Promises don't work well with WASM, I think.
- Raymond: The convolver was priority-2 so same for this.
- Paul: What do the attendees think?
- Jeff: I definitely prefer promises.
- Christoph: You can create a promise if panner has an onchange
event. But you can't do that with a promise.
- Paul: It's not uncommon for apis to have both.
- Christoph: Can you set HRTF in the constructor?
- Paul: Yes. Then a promise won't be delivered, but an event
would work.
- Raymond: I think in Chrome, constructing with HRTF is async,
so you'll get silence until it's ready. This is different from the
convolver. We could change it so it's synchronous, but
that seems not nice
when it's already async.
- Christoph: Can we deprecate the buffer and HRTF in the
constructor?
- Paul: Can't remove these. Best we can do is notify the user.
- All: A promise seems best.
- [Paul updates issue]
- AOB
- We're down to 10 V1 issues. (2 are admin issues.)
- US goes back to standard time and Europe was last week (?) so
the time difference will be back to the usual.
Thanks everyone, and we'll see you next week!
On Wed, Oct 28, 2020 at 2:16 PM Raymond Toy <rtoy@google.com> wrote:
> Here's the agenda for our WG/CG meeting.
>
> Meeting venue https://meet.google.com/wgg-wjtd-kdu
>
> Agenda:
>
> - V1
> - New issues
> <https://github.com/WebAudio/web-audio-api/issues?q=is%3Aopen+is%3Aissue+label%3A%22Needs+WG+review%22>
> - Audio worklet error handling
> <https://github.com/WebAudio/web-audio-api/issues/2267>
> - Reopened: Exclusive access to audio HW
> <https://github.com/WebAudio/web-audio-api/issues/72>
> - Status of outstanding issues that we should fix for v1.
> - Frozen arrays for AudioWorklet
> <https://github.com/WebAudio/web-audio-api/issues/1933>
> - ABSN.start control messages
> <https://github.com/WebAudio/web-audio-api/issues/2248>
> - Any others?
> - Open PRs to be reviewed
> - ABSN extrapolation
> <https://github.com/WebAudio/web-audio-api/pull/2256>
> - V2
> - New issues
> - PannerNode HRTF
> <https://github.com/WebAudio/web-audio-api-v2/issues/102>
> - New comments
> - NoiseGenerator
> <https://github.com/WebAudio/web-audio-api-v2/issues/8>
> - Incremental delivery of audio data for OfflineAudioContext
> <https://github.com/WebAudio/web-audio-api-v2/issues/66>
> - Remove compileStreaming
> <https://github.com/WebAudio/web-audio-api-v2/issues/79>
> - Multithreading
> <https://github.com/WebAudio/web-audio-api-v2/issues/85>
> - ConvolverNode multiple channels
> <https://github.com/WebAudio/web-audio-api-v2/issues/20>
> - priority-1 issues
> <https://github.com/WebAudio/web-audio-api-v2/issues?q=is%3Aopen+is%3Aissue+label%3Apriority-1>
> - AOB
>
>
Received on Thursday, 29 October 2020 17:30:08 UTC