See also: IRC log
<trackbot> Date: 21 November 2014
<joe> trackbot, end telcon
<trackbot> Meeting: Audio Working Group Teleconference
<trackbot> Date: 21 November 2014
<shepazu> joe, working on it
<joe> shepazu, I think I just got it with my explicit correction to zakim?
<cwilso> I'm running a few minutes late
<padenot> Clermont-Ferrand
<joe> cwilso, what's your eta?
<cwilso> I'm here
Bug review; starting with #72 (leaving #358 aside for now)
<ChrisL> http://msdn.microsoft.com/en-us/library/windows/desktop/dd370844%28v=vs.85%29.aspx
Cwilso notes that some OSes have lower latency if you have exclusive access (but has its own host of issues) - relates to this proposal.
<joe> q
<joe> ack
shepazu notes that he believes this relates to accessibility
<ChrisL> is ultra low latency a primary driver here? would someone use web audio while live tracking for example?
also needs to be some overall control (permission, revocation)
<shepazu> (this is a potential accessibility feature or problem, definitely need to ask permission, and a way to revoke permission)
padenot notes that interrupting other sources is a side effect - also notes that WSAPI exclusive access is actually buggy/unusable - doesn't think we should put this in v1
<padenot> cwilso: we do, but it's not really worked on in the audio wg
ChrisL thinks access to multiple devices is key, but other things are optional
<Zakim> cwilso, you wanted to say I thought we had a separate issue for output device selection; we should.
<cwilso> https://github.com/WebAudio/web-audio-api/issues/445
Cwilso notes that while webrtc has a device selection API, but Web Audio doesn't have a way to indicate which context has filed placeholder issue for this; OK with deferring #72 from v1.
On to #49
cwilso thinks that letting user choose between a-rate and k-rate is a bad idea; asserts that there is nothing to do here however AudioWorkers may require more discussion
<ChrisL> +1 to close
Note to close
On to #45
Joe notes there is lots of developer-oriented documentation that isn't even informative
padenot agrees, though without specific examples
shepazu proposes Matt and he walk through the spec and red-line portions that are suitable for extraction as developer documentation source data
+1 on this...
<joe> ACTION: Doug and Matt will identify sections of spec to be broken out as per issue 45 [recorded in http://www.w3.org/2014/11/21-audio-minutes.html#action01]
<trackbot> Created ACTION-116 - And matt will identify sections of spec to be broken out as per issue 45 [on Doug Schepers - due 2014-11-28].
On to #44
cwilso is good with this.
billhofmann notes that 32 is arbitrary
padenot notes that you need a limit, but 32 is too low - he has use cases
chrisl asks why limit?
<ChrisL> aha. at least is a lower limit, not an upper
so, the real question (per cwilso) is that this really should be "what is the *minimum* number of channels"?
Per cwilso - the question really is whether an implementation fails not having 128 channels, is that non-conformant?
Cwilso - would an implementation that explicitly limits be non-conformant?
Joe suggests we leave it as is for now (32 channel floor)
Joe says - we need to state a little more consistently through the spec perhaps?
Cwilso says that he doesn't think this is an issue, really - it's used elsewhere where relevant
Joe says - everything is ok, we can close this issue; cwilso and padenot agree that a floor is sufficient.
On to #16
Cwilso notes that this is a common request, including in webmidi
cwilso also notes this is a reasonable request; somewhat difficult to implement however
Joe proposes keeping this in the v1 bin (to be addressed again later if there is major concern)
cwilso agrees
(someone is dancing on their keyboard)
On to #9
Joe notes that this is a nice-to-have, you can always create a new node
cwilso thinks this might be higher than nice-to-have, but needs more thought
Joe proposes that a case is better made at a node-per-node level
Cwilso agrees, but thinks AudioWorker might be a specific case
shepazu thinks that a per-node reset case is befuddling.
Joe notes that making it general brings up a bunch of new issues related to state, etc - e.g., can you now call start and give it a different time?
Joe says we need more time on this, needs more exploration
Joe says we'd need to walk through all nodes and understand this
cwilso agrees
billhofmann asks whether the lack of strong feeling suggests we move on
Joe moves to v.next bin
On to #7
Cwilso and Padenot say, "close!"
Joe notes this really requires a more complete discussion.
Nine issues left, now 8 minutes left.
Joe proposes we adjourn early and come back after Turkey Day
Joe suggests next meeting date is Dec 4
<cwilso> I got rid of two of the 9 issues (they were set to v.next, so we didn't need to review)
billhofmann notes that Dolby is considering whether we can contribute on the testing effort
Joe notes we should have that discussion on the list
<joe> trackbot, end telcon
This is scribe.perl Revision: 1.140 of Date: 2014-11-06 18:16:30 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/better behavior/lower latency/ Succeeded: s/ipper/upper/ Succeeded: s/sense/request/ No ScribeNick specified. Guessing ScribeNick: BillHofmann Inferring Scribes: BillHofmann WARNING: No "Topic:" lines found. Default Present: +1.617.455.aaaa, joe, +1.510.387.aabb, Doug_Schepers, padenot, ChrisL, BillHofmann, rtoyg, hongchan Present: +1.617.455.aaaa joe +1.510.387.aabb Doug_Schepers padenot ChrisL BillHofmann rtoyg hongchan Found Date: 21 Nov 2014 Guessing minutes URL: http://www.w3.org/2014/11/21-audio-minutes.html People with action items: doug identify matt of sections spec will WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]