Re: [Agenda] WG/CG Telecon Thurs Jul 8, 9:00 am PDT

Attendees:  Jeff Switzer, Paul Adenot, Christoph Guttandin, Raymond Toy

Minutes:

   - Admin
      - Review of process to update the spec
      <https://github.com/WebAudio/web-audio-api/issues/2362>
         - [Raymond explains the key points of the process]
      - V1
      - PRs
         - Fix #2339: Fix ecma262 completion
         <https://github.com/WebAudio/web-audio-api/pull/2350>
            - Raymond: Just needs a quick review; no need to add amendment
            boxes since this is already in the recommendation
         - Use new Web IDL buffer primitives
         <https://github.com/WebAudio/web-audio-api/pull/2363>
            - [Raymond describes how the amendment boxes work]
            - Paul: This looks pretty good; a list of changes for the
            change log is good, but too bad it doesn't show up in the preview.
            - Raymond: Yeah, a bikeshed problem where bikeshed processes
            the script tag as regular text.  Could possibly be worked around.
         - Fix typo in VUMeterNode
         <https://github.com/WebAudio/web-audio-api/pull/2356>
            - Raymond: Simple typo fix, but needs to land after the WebIDL
            one, since that has all the infrastructure in place.
         - RangeError for only for negative cancelTime
         <https://github.com/WebAudio/web-audio-api/pull/2352>
            - Raymond: Must land after WebIDL to get the infrastructure
            changes.
         - V2
      - 16 Priority-1 issues
      <https://github.com/WebAudio/web-audio-api-v2/issues?q=is%3Aopen+is%3Aissue+label%3Apriority-1>
         - user-selectable size
            - Raymond: As discussed last week, we can proceed with writing
            a PR for this.  One issue is the size of the
SciriptProcessor buffers.
            Still power of two?  The proposal says power of two times
the render size.
            - Paul: Still has to buffer, so power of two and not doesn't
            make much difference.
            - Paul: Maybe need to add that convolvers may add additional
            latency?
            - Raymond: Chrome uses PFFFT so powers of 2,3,5 are supported
            so no additional latency needed.  And browsers are allowed
to round up to
            the next supported size.
            - Paul: That's complicated on windows with 441 or 480.
            - Raymond: Neither of these are supported by PFFFT since 441 is
            powers of 3 and 7.  440 (which chrome uses) is a power of
2, 3, and 11.
            480 is power of 2,3,5, so supported.  Could extend PFFFT
to support 7 and
            11, but it will be slower.  But browser could round up
440/441 to 480.
            Might be ok?
            - Paul: Yeah, it's pretty complicated.
         - output selection
            - Paul: Well under way in Firefox
            - Raymond: Just adding a sinkID to the constructor dictionary?
            - Paul: Yes.
         - AOB


On Thu, Jul 8, 2021 at 8:33 AM 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:
>
>    - Admin
>       - Review of process to update the spec
>       <https://github.com/WebAudio/web-audio-api/issues/2362>
>    - V1
>       - PRs
>          - Fix typo in VUMeterNode
>          <https://github.com/WebAudio/web-audio-api/pull/2356>
>          - Fix #2339: Fix ecma262 completion
>          <https://github.com/WebAudio/web-audio-api/pull/2350>
>          - Use new Web IDL buffer primitives
>          <https://github.com/WebAudio/web-audio-api/pull/2363>
>          - RangeError for only for negative cancelTime
>          <https://github.com/WebAudio/web-audio-api/pull/2352>
>       - V2
>       - 16 Priority-1 issues
>       <https://github.com/WebAudio/web-audio-api-v2/issues?q=is%3Aopen+is%3Aissue+label%3Apriority-1>
>    - AOB
>
>
>