See also: IRC log
<trackbot> Date: 10 July 2014
<cwilso_> ahh.
<cwilso_> so I see.
<jernoble> well, i guess i won't tell it who i am, then. :)
<kawai> Zakim is down mean it is not possible to call?? # sorry for really basic question...
<padenot> kawai: the bridge works, but the irc bot is down
https://www.w3.org/2011/audio/track/agenda
<cwilso> kawai: no, you can still call in to the telecon, just no attendee/agenda management
ACTION-104
<trackbot> ACTION-104 -- Paul Adenot to Extract formulas from cookbook -- due 2014-06-26 -- OPEN
<trackbot> http://www.w3.org/2011/audio/track/actions/104
<kawai> cwilso: Thanks!
padenot: no progress
ACTION-104 due in 2 weeks
<trackbot> Set ACTION-104 Extract formulas from cookbook due date to 2014-07-24.
<shepazu> http://shepazu.github.io/Audio-EQ-Cookbook/audio-eq-cookbook.html
action-105
<trackbot> action-105 -- Doug Schepers to Mark up formulas for biquadfilter in mathml (get help from mathjax?) -- due 2014-06-26 -- OPEN
<trackbot> http://www.w3.org/2011/audio/track/actions/105
shepazu: collaborating with
RBJ
... ASCII math in the original cookbook was a bit odd at times,
still working on it
... been working with Peter Krautzberger from Mathjax
... converting ascii math to mathml and using mathjax to
render
... also working on adding structure to make it easier to
follow
... RBJ also had suggested changes
... tone is very informal, which may not be best for spec
padenot: prefer when docs are in the same tone
shepazu: i think we should give instructions on what to do rather than suggest
padenot: think that's fine
shepazu: second question is whether this will be published as separate doc or part of webaudio spec, as appendix or otherwise
padenot: pannernode has all
equations inline, which is handy for implementors
... best to keep eqs inline, maybe refer to the full
document
shepazu: we could do both - as a
WG we could publish this doc, cleaned up, as a math overview,
informative document (a note)
... as well as extract the math for the spec
olivier: why publish separately if we can have stuff inline and add RBJ to acknowledgments?
shepazu: give the doc a permanent home, something to reference
olivier: good enough rationale
shepazu: it's also a better way
to give full credit to the author
... 3rd question about mathml (and mathjax for some
browsers)
... would be the second spec using mathml
... and the first non-mathml spec using mathml
... propose to require single version of library centrally in
w3c space
padenot: we should have the lib in the repo, for offline use
shepazu: might also want to offer
pdf version of the spec
... alternative was to have image fallback, not sure we liked
it
cwilso: slight pref for not
having pdf version unless necessary
... not convinced really needed
shepazu: will bring that back to
plh
... we have traditionally offered various versions of
specs
... also need to figure out what we want to do with references
to openal
olivier: I failed to find authoritative spec, only implementation code
shepazu: if that's the case we may want to remove references, let me chase
<rtoyg_> Is this the spec: http://www.openal.org/documentation/openal-1.1-specification.pdf
action-105 due in 2 weeks
<trackbot> Set action-105 Mark up formulas for biquadfilter in mathml (get help from mathjax?) due date to 2014-07-24.
action-97 DONE
action-97: DONE
<trackbot> Notes added to action-97 Start spreadsheet to triage issues for webaudio api.
action-97 CLOSED
<trackbot> Closed action-97.
https://github.com/WebAudio/web-audio-api/issues
olivier: top priority -
https://github.com/WebAudio/web-audio-api/issues?labels=Architectural%2FFundamental+%28Breaking+change%29&page=1&state=open
... tagging will influence priorities, we will start with the
"architecture" and the "clarifications (require changes)"
tags
... if you disagree with tagging, say so as it will determine
issues we start with
https://github.com/WebAudio/web-audio-api/issues/113
olivier: don't see any discussion since may, what do we need?
cwilso: have been wanting to
prototype
... and consider the idea of where latency is produced and how
we can eliminate it
... and then used SPN as a conceptual building block of the
rest
... problem is we double-buffer the node
... we have to jump across a thread boundary
... talked with Ray about pushing audio processing into the
audio thread
... run SPN in the audio thread, with possibility of pushing to
separate thread if needed
... create normal worker threads as needed, introduce yr own
latency
... question of how it aligns with current
implementations
... personal preference of removing ability to build on the
main thread
padenot: agreed - we can't keep
the current situation
... been spending time trying to make it better, made some
progress but not sustainable in long term
... been looking at demos, people porting stuff, not easy to
just express with webaudio api
... and upset when it doesn't work with SPN
... been talking months ago with ROC on how to do that
cwilso: one thing not captured in
code in the issue thread is to have the event handler not be an
event handler
... use sync callback instead
padenot: we talked about shader
model
... you upload your shader, no change
cwilso: main thing would be to
remove as much async as possible
... spoke to vendors
... can't arbitrarily split graph between 2 threads
padenot: ideally you should be
able to re-implement the api with it
... perhaps not with same performance but should be an
aim
... at least a good exercise
... happy to talk at implementation level
... involve ROC, Ehsan
cwilso: would be good - should take action item away
<scribe> ACTION: cwilso to document synchronous option in github issue 113 [recorded in http://www.w3.org/2014/07/10-audio-minutes.html#action01]
<trackbot> Created ACTION-106 - Document synchronous option in github issue 113 [on Chris Wilson - due 2014-07-17].
olivier: probably top agenda for potential implementors f2f
cwilso: still trying to organise
that
... music/audio hack day in Berlin, might be an option
... main thing I want a solid feel from rest of the group
... preference for removing current SPN
... will be a big breaking change
... we would have to have good documentation of the
transition
... including knowledge of workers
olivier: sounds like candidate
for CfC
... could have it now. decoupled from rest of the
discussion?
<scribe> ACTION: olivier to organise CfC on breaking change for the main thread scriptprocessornode [recorded in http://www.w3.org/2014/07/10-audio-minutes.html#action02]
<trackbot> Created ACTION-107 - Organise cfc on breaking change for the main thread scriptprocessornode [on Olivier Thereaux - due 2014-07-17].
https://www.w3.org/2002/09/wbs/35125/TPAC2014/
olivier: please register ASAP or contact us if you can't make it
shepazu: please attend - important meeting
<cwilso> http://channel9.msdn.com/Events/Build/2014/3-548
suggested: 31st July
cwilso: possible regrets
This is scribe.perl Revision: 1.138 of Date: 2013-04-25 13:59:11 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/Peter/Peter Krautzberger/ No ScribeNick specified. Guessing ScribeNick: olivier Inferring Scribes: olivier Present: cwilso kawai rtoyg_ shepazu olivier jer Agenda: http://lists.w3.org/Archives/Public/public-audio/2014JulSep/0021.html Found Date: 10 Jul 2014 Guessing minutes URL: http://www.w3.org/2014/07/10-audio-minutes.html People with action items: cwilso olivier[End of scribe.perl diagnostic output]