See also: IRC log
Clarke: let's get started, agenda. First item is look at Jan's comments
Jan: there is one more change; split one error..is that ok?
Clark: Discuss Separating Auth errors from other
Mark: finishing off proposal, work on messaging and will post on wiki in next couple days
<kaz> Jan's write-up
Clarke: next item is from Jan on OIPF
Jan: There are 3 documents that
are referenced...
... 1 - document about architecture
... they use Marlin ...
Clarke: I've put up pages on Adaptive bitrate
<ph> http://www.w3.org/2011/webtv/wiki/MPTF/ADR_Error_Codes
<ph> http://www.w3.org/2011/webtv/wiki/MPTF/Content_Protection_Error_Codes
mav: What is the IP around using
OIPF?
... What is the IP around using OIPF?
Jan: there is precedence with a license
Clarke: what is procedure/process for referencing outside documents?
Philipp: We should use the document?
mav: no, we should use ideas and
incorporate them...
... no, we should use ideas and incorporate them...
ph: We need to be sure we're interested enough to use doc/ideas
Clarke: some preliminary investigation would be good
<kaz> W3C Patent Policy
glenn: why did we wish to incorporate ideas, instead of reference sections in OIPF?
<kaz> ACTION: hoschka to look at IP issues with referencing OIPF documents [recorded in http://www.w3.org/2011/12/08-webtv-minutes.html#action02]
<trackbot> Created ACTION-83 - Look at IP issues with referencing OIPF documents [on Philipp Hoschka - due 2011-12-15].
jan: There is a precedence with OMA
juhani: section on patent policy that are referenced
kaz: we should clarify what we want to refer to/include
Clark: Juhani work with Jan to get guidance and clarify what will be included
<Juhani> "Juhani work with Jan to get guidance and clarify what will be included" - I think this is someone else than Juhani this action was appointed to ...
<glenn> glenn: notes that section 8.2 of PP allows exclusion of claims based on "mere incorporation by reference"
<kaz> ACTION: hoschka to work with Jan to get guidance and clarify what will be included [recorded in http://www.w3.org/2011/12/08-webtv-minutes.html#action03]
<trackbot> Created ACTION-84 - Work with Jan to get guidance and clarify what will be included [on Philipp Hoschka - due 2011-12-15].
Clarke: next agenda item, WHATWG...around adaptive bitrate proposal
<Clarke> Adaptive delivery page: http://www.w3.org/2011/webtv/wiki/MPTF/ADR_Error_Codes
Clarke: also created page to summarize this
Duncan: don't read too much into the source
<ph> page should probably link to source in whatwg
Clarke: organized into sections around params, error codes, feedback, new apis
Duncan: would be useful to have
'bytesReceived' under feedback
... bytes received is more explicit
Clarke: lets look at items.. parameter, maxLevel, it is the bitrate level
Duncan: This proposal is moving towards remote control approach that JS can influence bitrate selection
MarkW: need lots of data and feedback in JS to enable heuristics to make good decisions for bitrate selection
Duncan: added seconds on next segment to help logic
<ph> link to original whatwg page?
MarkW: it's unclear that we have enough information and details in proposal
Clarke: we should discuss the 3 architectures/modes
MarkW: concern over mode three about sending video data through JS
<ph> whatwg source seems to be http://wiki.whatwg.org/wiki/Adaptive_Streaming
MarkW: the google approach does use AJAX to retrieve data
Bob: there is same issue that useragent needs to expose metrics regardless
MarkW: approach 3 is easier, but approach 2 the useragent is aware of bitrates and manifests
mav: video append would be useful for smooth playlists or inserting ads
<ph> +1 on mav
<mark> +q
kaz: we should include the original resource of WHATWG working group as Philipp pointed out, shouldn't we?
<kaz> http://wiki.whatwg.org/wiki/Adaptive_Streaming
<duncanr> and http://wiki.whatwg.org/wiki/Video_Metrics
jason: is the 'currentSegmentDownload' a best guess of seconds based percentage? or real seconds?
Duncan: best guess based on average bitrate
Clarke: deadline of Dec 15 for
HTML5
... we should work offline to get our proposal tight and use
wikipage to send recommendations
MarkW: unlikely that timeframe is
possible
... benefit of approach 3 is enable experimentation....
... only add details to approach 1 to just report bitrate
selected
Clarke: we could get everything into html5, or try to influence implementations that drive back to standards
Duncan: setting the level valuable, and querying the available levels
Kilroy: there need to be ability
to set max bitrate ceiling, but beyond that requires more event
feedback to enable heuristics
... at early state of adaptive streaming, enabling apporach 3
(all in JS) gives advantages
MarkW: unlikely HTML5 group may accept
<ph> proposed resolution: go for option 1
<mav> Kilroy, do you think option 3, as you described, can be proposed by 12/15?
MarkW: timing of socializing this is issue
Clarke: let's continue the discussion using the mailing list and the wiki by Tuesday
<mav> What exactly has to be complete by 12/15? Does the proposal need to be finalized? Socialized?
<kaz> mav and Clarke, maybe we should check with Paul Cotton as well
[ adjourned ]