Media Pipeline Task Force Teleconf

21 Jun 2012



Clarke, Dave_Mays, duncanr, johnsim-microsoft, acolwell, nisc, Bob, ddorwin, Kazuyuki


Bug Review

Clarke: checked yesterday, no updates on bugs
... latest was mid to end of May

review changes to Content Protection requirements

<Clarke> CP Requirements: http://dvcs.w3.org/hg/webtv/raw-file/tip/mpreq/MPTF-CP-Requirements.html

Clarke: finished filling out the draft doc, probably needs some cleanup
... changes are in the use cases
... everyone please read through in the next week and provide comments back
... once this is done, MPTF work should be on-demand
... we should be able to disband soon
... reviewing UC 4.2.2
... added implementation, motivation and dependencies

<ddorwin> do you expect to switch in the same stream?

<ddorwin> without a load()?

<ddorwin> switch content protection systems in the stream?

<ddorwin> that would be very problematic for implementations

Johnsim: common thing would be going from protected to unprotected and back

Clarke: less likely, but also possible, ads could be protected with a different scheme than the primary content
... moving on to 4.2.3

<ddorwin> This assumes content is encoded in a CDM-specific way, which we want to avoid.

<ddorwin> (bullet 2)

Johnsim: not possible to have open-source DRM

<ddorwin> Is this referring to something like Clear Key (in the EME proposal)?

Johnsim: is the requirement meant to cause a UA to expose a third-party CDM, rather than providing an implementation?
... if UA is implementing ClearKey, that's fine
... we should amend the requirement to make it more clear and unambiguous

Clarke: no problem specifying Clearkey here. That will make it easier to understand

davidmays: please expand CDM with a definition

Clarke: moving on to 4.2.4

<ddorwin> I think that last sentence you said is worth including in this section.

Johnsim: what do you mean by credentials here?
... there are lots of different ways credentials get communicated

Johnsim: is this requirement implying that the browser is required to store literal login/password credentials

Clarke: tried to make it clear in bullet 3
... should not be limited by the browser

davidmays: maybe replace "credentials" with "authorization assertion"

Johnsim: agreeing with authorization assertion

Clarke: I'll put that in there
... moving on to 4.2.5
... this is just here to make sure it works for both adaptive and non-adaptive content

davidmays: i think the term "copy protection system" is a little too specific

Clarke: ok I will change that
... we will turn this in, and then we will basically be done

<Clarke> Thanks, David,

no problem


<kaz> [ adjourned ]

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.128 (CVS log)
$Date: 2012/06/21 16:17:42 $