W3C

Web and TV IG: GGIE meeting

9 Sep 2015

Attendees

Present
Glenn Deen, Bill Rose, Nilo Mitra, Kazuyuki Ashimura, Paul Higgs, Leslie Daigle, Bill Hoffmann, Giri Mandyam
Regrets
Daniel Davis
Chair
Glenn
Scribe
Bill

Contents


Call admin

<glennd> The remaining online meetings will be to prepare for the Japan F2F. The tasks are to review all Use Cases; Identify which ones should be eliminated or postponed until 2016; identify which are important for this year but which may need additional discussion.

Review of aggregated set of use-cases that have been submitted during the past 8 months.

<glennd> Looking at the GGIE Common Requirements figure from the wiki (Smart Edge-Creation/Core/Smart Edge-Viewing) https://www.w3.org/2011/webtv/wiki/GGIE_TF I’d like to create a large poster of the figure for TPAC. If anyone has any changes or comments on the figure please let me know.

<gmandyam> When we discuss the UCs should the Core include broadcaster requirements/issues on supplemental services such as emergency alerting?

<glennd> Good question. We have not addressed that.

Action Item: Capture question for further discussion, perhaps for 2016.

Discussion on User Discovery UCs

<glennd> https://www.w3.org/2011/webtv/wiki/GGIE_TF/UseCases/User_Discovery

<glennd> There has been some confusion as to whether this means discovering the user versus discovering content. Is there any discussion on the 3 User Discovery UCs? Should we rename to avoid confusion?

<Bill Hoffmann> Should we rename to avoid confusion?

<glennd> Good idea.

Action Item: Bill Rose to rename the category from “User Discovery” to “User Content Discovery”

<Bill Hoffmann> Is guided discovery included in this category? E.g. “if you like that, you might like this”

<glennd> We did not discuss guided discovery. Glenn asked Bill to write up a UC on it.

<gmandyam> Cannot assume things like “last channel” is known when discovery crosses delivery channels

Action Item: Bill H, Giri will work on UCs to cover guided channel, cross delivery channel discovery.

Discussion on Viewing UCs

<glennd> https://www.w3.org/2011/webtv/wiki/GGIE_TF/UseCases/Streaming

<gmandyam> Could not find anything covering interactivity, timed media events, data queues, etc.

<glennd> Good point. Is there some related work by the Second Screen group?

<Kaz> I posted a URI from the Second Screen WG with 6 UCs. https://github.com/w3c/presentation-api/blob/gh-pages/uc-req.md

<gmandyam> I was not really thinking about second screen for this. Will post some ideas to the reflector. Should this be an extension of UC-1 or UC-2?

<glennd> I think it is UC-2 Streamed Content Viewing

Action Item: Giri will add a UC covering interactivity, etc., based on Viewing UC-2 Streamed Content Viewing.

<glennd> Streaming UC-3 Identifying Streamed Content covers the concept of an application identifying the content it is playing and using the identification information to obtain a list of server URLs for the content, select a URL it wishes to access the content from, record it, perform trick play, etc. Some clients have some Automatic Content Recognition built in but not specifically for this purpose (mainly ads). SMPTE is working on adding IDs (AdID, EIDR) to content. This UC goes substantially beyond what is being done today.

<glennd> Streaming-UC-4 Manipulating streamed content covers the situation where content items may be interposed with other content to create the version that is streamed to the user. E.g. ad insertion.

<Nilo> This is the top level UC for e.g. ad insertion

<glennd> Streaming-UC-5 Streamed Content Ad Insertion Interstitial Ads. Covers dynamic insertion of ads from secondary parties. Are there any other related UCs for streaming we are missing? There were no responses.

Discussion on Content Identification UCs

<glennd> https://www.w3.org/2011/webtv/wiki/GGIE_TF/UseCases/Content_Identification

<glennd> UC-1 User Device Retrieval of Content Address(es) Using EPG/Title covers how to resolve an address from the EPG title which includes the content ID. Are there any additional Identification or measurement UCs that should be added?

<Kaz> There are opportunities for demos and breakout sessions at TPAC. Can we show any demos on GGIE?

<glennd> Lets take this up on the reflector.

<Leslie> Are you suggesting breakout sessions are different from posters or demos?

<Kaz> We can do a poster to bring people to a breakout session on Wednesday after the Plenary to broaden audience for GGIE.

<Leslie> Are you then suggesting we hold a poster session and followed up with some breakout talks?

<Kaz> Yes. Wednesday would gain a wider audience than Monday.

<glennd> Let's do it.

Action Item: Glenn to work with Kaz to add a breakout session. Also to look at who is registered from GGIE for TPAC and to get one or two to speak.

Discussion on Source Access UCs

<glennd> https://www.w3.org/2011/webtv/wiki/GGIE_TF/UseCases/Source_Access

<glennd> This category basically focused on caching. Source-Query-UC-1 Basic Query by Content identifier URI addresses the user querying a lookup service using a content URI which finds the authoritative service which returns the results through the lookup service to the user application.

<gmandyam> Does the source of the identifier matter? Do we need to address e.g. the client support for a watermark detection/resolution mechanism at the client to accomplish this?

<glennd> This UC does not address where the URI comes from. The app somehow “knows” the URI for this UC. UC-2 “Query by Content identifier URI with delegated sub-namespace” gets into how to resolve the URI via a hierarchy of services.

Discussion on Cache UC-1 Cache Query and Cache UC-2 Cache Discovery (same page as Source Access)

<glennd> UC-1 covers how to query a cache for content and the content original source. UC-2 covers a cache advertising itself on the network so it can be discovered by the viewing device. We may need more discussion on this subject before TPAC.

Discussion on Content Capture UCs

<glennd> https://www.w3.org/2011/webtv/wiki/GGIE_TF/UseCases/Content_Capture

<glennd> Capture-UC-2 Assigning a unique content identifier at asset capture gets into how a capture device obtains a unique content identifier. Capture UC-3 Fingerprinting and associating a content identifier at asset capture covers the capturing device generating a fingerprint of the captured content and registering it along with the unique content ID.

<glennd> Capture UC-4 3rd Party Composite Asset From Multiple Sources. Allows a 3rd party to create a new asset from one or more original assets (commercial, social, etc.), potentially adding new content e.g. audio, and making the new asset available. UC-5 adds to UC-4 by allowing the new asset to be sent as a series of links to the original assets along with timing information, etc., necessary to allow the client/viewer to assemble the asset from the original sources. This is a very sophisticated UC. Should we write up a narrative for this UC to illuminate it better? This might be a good one to discuss at TPAC at a breakout session as a next gen video challenge.

Action Item: Bill to create a narrative for this UC and send to Glenn for discussion.

<glennd> Are there any other Capture scenarios we need to add for the 2015 cycle? There were none. Getting back to the big picture, we have not covered User Identity, privacy issues, or ingest this year. Are there any other issues we have not identified? [None were noted.]

Next meeting

<glennd> Sept 23; October 7, both at 11:00 AM ET. Note: October 7 will be the final call before TPAC. No call on October 21.

<glennd> Meeting adjourned

Summary of Action Items

Action Items created during this call

[End of minutes]