Re: [GGIE] Draft response to questions

Hi Glenn,

Thank a bunch for writing the draft response and reflecting the TF members'
suggestions we had during the call last week.

My personal comments (I mean, my co-chair hat off) are in line below.

On Fri, Jun 19, 2015 at 3:01 AM, Deen, Glenn (NBCUniversal) <
glenn.deen@nbcuni.com> wrote:
[snip]

>  -------------------------
>
> Here is a rough English translation of the comments I've gathered during
> the conversation with engineers, researchers and media strategists from NHK
> [1] and JBA [2] on the GGIE use cases. Note that the conversation was in
> Japanese and really casual, so this English translationmight contain
> something beyond, or different from, their original ideas. If you found
> something wrong or logically invalid in these comments, that would be my
> bad.
>
>
>
> I've already shared with the moderator and co-chairs these comments; Glenn
> is preparing his responses, and it's likely for us to discuss the comments
> and the responses at the next call or sometime in the near future.
>
>
>
> Regards,
>
> Yosuke
>
>
>
> [1] http://www.nhk.or.jp/corporateinfo/
>
> [2] http://www.j-ba.or.jp/category/english
>
>
>
>
>
> --------
>
> ** Standardization Process*
>
> How will succeeding works in other SDOs happen?: The standardization
> process after we finish analyzing gaps in the TF is unclear, esp.
> succeeding standard developing works in other SDOs such as IETF and SMPTE:
> Is Glenn going to bring the document privately to them, or will the W3C Web
> and TV IG GGIE TF make liaisons with other relevant SDOs on the gap
> documents and encourage them to start developing new standards?
>
>
>
> *The GGIE group plans on delivering GAP analysis identifying where there
> are missing features in standards and which SDO would be in a position to
> work on those features.   To help the SDOs we will also create prototype
> charters for the efforts describing the features we are asking them to
> consider working on.   These will be delivered to the Web & TV Interest
> Group during the W3C TPAC meeting in October in Sapporo*
>
>
>
> *SDOs vary in how they accept new work proposals and we expect that we
> will have to tailor each proposal to meet the specific SDO to which it is
> being submitted.  We will use formal liaisons to share the output of the
> group to each SDO, and in addition where it is appropriate members who
> support the GGIE proposal and want to work on it in can also *
>
> *engage the SDO and its community to create support for the work item. *
>

Additionally, I recommend we contact and try to involve experts at other
relevant SDOs when we do the gap analysis.  It'd help us better the
analysis and raise the awareness of the experts, who might lead succeeding
works in the SDOs.


>
>
>
>
> Involving streaming media stakeholders: Should we invite streaming media
> service providers such as Youtube (Google) and Netflix intothis discussion?
> They seem like important stakeholders in W3C on this type of topics.
>
>
>
> *That does seem like a logical idea.  Everyone is very welcome.*
>

We've agreed that Daniel, one of the team contacts of the IG, would contact
the AC reps of Google and Netflix to encourage them to join this
conversation.


>
>
>
>
>
>
>
> Overlaps with TV-Anytime: There are some overlaps between the
> concept/scope/architecture of GGIE and that of TV-Anytime: CRID <content
> reference ID> , user metadata, authority, resolution providers, etc.
> Jean-Pierre from EBU has already mentioned about some of them but the
> discussion seemingly stopped before reaching any conclusion in the TF. Is
> the TF going to further look into TVA? [1]
>
>
>
> *GGIE is looking at end-end digital video on the Internet.   This includes
> looking at what is done today through use-cases specifically capturing
> current practices and architectures. It also includes looking at use-cases
> that capture what we anticipate will be future features that are not
> available today.   Through understanding today’s technology and
> understanding tomorrow’s use-cases we hope to identify work to be done in
> SDOs to help bridge today’s available solutions into tomorrow’s wanted
> systems, without throwing away what is done today.   To put it another way,
> the intent is to evolve the ecosystem not reinvent it, and to make the new
> evolved features available via open standards.*
>
>
>
> *GGIE is limited in it’s scope to use-cases and not specific technology
> selection.   For the use-cases reviewed it is expected that there are many
> technology solutions to choose from. We will attempt to identify the common
> features across the technologies that enable the use-cases to be
> implemented and also identify through the GAP what common useful features
> are not available via standardsbased specifications today.*
>

I for one also see some overlaps between the discussion we have in the TF
currently and the discussion the TV-Anytime forum had about ten years ago,
which was when I contributed to the TVA2.0 work. I'm not saying we should
adopt or select TVA but just suggesting that if we have TVA experts in our
discussion it would help save some of our time thanks to their experience
in overlapping areas.

I'll try to involve them more, esp. on specific use cases or topics on
which we might get benefits from their expertise.


>
>
>
>
>
>
> ** General Comments on Use Cases*
>
> - Boundary of the ecosystem: What will be the boundarybetween the services
> which implement, or is a part of, the ecosystem and the rest of the world?
> In other words, what is the scene setting about the boundary behind the use
> cases? Should we review the use cases with assuming the world where all
> video content distributions on the Net follow the GGIE use cases?
>
>
>
> *The focus for GGIE is on the entire end-end (glass-glass) ecosystem for
> digital media on the Internet.   That does make the scope very large.
> However, there is a lot that is out of scope; for instance GGIE will not
> include features of particular steps in the end-end process which are local
> only to those steps.  Put another way, GGIE will not work on use-cases on
> how editing of video is done, except where there is an impact to the larger
> end-end workflow.   More spherically GGIE would not focus on dubbing/mixing
> workflows, but would examine how a unique content identifier would enable
> integrating the editing cycle with capture and later distribution of the
> edited work.*
>
>
>
> *GGIE’s scope is about movements of digital media data and metadata from
> end to end.*
>

Let me check my understanding by an example.

HTTPS is an end-to-end encryption protocol which doesn't form any
restricted/privileged groups or area on the Net, while some VPNs on the Net
use end-to-end encryption protocol(s) which form secure groups of computers
hence has a boundary between a virtual private network and the rest of the
world.

You mean GGIE is end-to-end like (not VPN but) HTTPS in the above example,
right?

This is my understanding based on the conversation we had in the call. I
might've misunderstood something.


>
>
>
>
>
>
>
>
>
>
>
>
> - How to involve non-professional users: Additionally, are there any ideas
> to make home or hobby video creators to follow the use cases when they
> upload or distribute content on the Net? Currently it seems like quite
> difficult to achieve it if we don't provide them with convenient
> mechanisms, incentive, and/or enforcement to do it.
>
>
>
> *This is a very good question.  One of the challenges is finding where
> non-professional users stop being non-professional.  On the net today
> non-professional users are creating content and using technology that a few
> years ago was exclusively used by professionals.   For that reason we think
> that the use cases we capture today which may include features that are for
> professionals will soon be relevant to non-professionals as they adopt the
> professional tools and abilities.  A related example is the increasing
> prevalence of monetized non-professional content, and the mechanisms that
> enable it such as content identifiers are essential to making that
> ecosystem work.*
>
>
>
> *One difference between professional and non-professional media creators
> is that professional creators will pay for things like content identifiers
> to be assigned to their works etc.   The non-professional is unlikely to
> directly pay for a content identifier, but they can obtain access to things
> like content identifiers by selecting an Internet based service that they
> use to upload and distribute their work.   So non-professionals get access
> to the features and ideas in GGIE not by directly buying them, but by
> accessing them from their distribution/social media service providers.*
>
>
>
> *GGIE is inclusive of both professional and non-professional content and
> we try to find common needs of both types of users.*
>

We might better contact the AC reps of Facebook and other social network
and/or media service providers in addition to Google and Netflix.

Daniel and Kaz, what do you think?


>
>
>
>
>
> Viewer reviews on video content: Viewer reviews can be considered as a
> type of metadata for video content (Cid) generated by users (Uid). Is it in
> the scope of GGIE?
>
>
>
> *User Identity and metadata data are in scope of GGIE but have not yet
> been examined in any detail so far.  We have only been active since January
> 2015 and have had time to initially focused on capture, search-discovery,
> and distribution.   User Identity and Privacy as well as more extended
> metadata are all in scope as we continue our discussions.*
>
>
>
>
>
>
>
> Terminal or household id: For the TV set in a living room, terminal or
> household id is sometimes more convenient, or better fits actual use cases,
> than user id. What does the TF think about this?
>
>
>
> *Increasingly personalization is becoming important to users and is
> something we need to examine how it can be accomplished and what the
> technical requirements and implications are for doing so.   GGIE will also
> have to consider privacy implication as it looks a the issue of identity of
> users.*
>

Let me check my understanding here again. Do you mean the TF exclude
terminal or household id, or something like family id/view, from its scope?

I just thought GGIE also covered the current primary use cases for people
to watch programs/videos on TV sets in some countries. Currently the use
cases mainly use broadcasting signals but which is likely or easily to be
replaced by the Net in the future.

Actually I think this would be a good decision if the use cases would
disappear in the near future. But I'm not so confident about it. I think if
we can get more inputs from the TF members on this topic, that'd help us
make the right decision.

Any thoughts on this item from other TF members?

Cheers,
Yosuke



>
>
>
>
> Anonymous users: Some use cases which prerequisites users to be identified
> might be more beneficial to the ecosystem if we allow anonymous users to
> use a whole or part of such use cases.
>
>
>
> *Privacy is essential in digital video consumption. Finding ways to enable
> it are important and part of GGIE.*
>
>
>
>
>
> * Comments on Specific Use Cases
>
> [User-Discovery-UC-1] Device profiles are definitely useful and looks like
> a missing part in the current Web standards. Security and privacy
> consideration should be done along the way.
>
>
>
> *Agreed.*
>
>
>
>
>
> [Streaming-UC-3] Does this use case include the case where a user feeds a
> stream from his or her home device to mobile devices?
>
>
>
> *It wasn’t designed with that specifically in mind, but yes it would cover
> that use.*
>
>
>
>
>
> [Streaming-UC-3] There is no description about to whomapplications report
> logs or measurements. Defining it as an actor might help better the use
> case.
>
>
>
> *Thank you for the suggestion. More future work needs to be done by the
> group on measurement, and we will incorporate that suggestion.*
>
>
>
>
>
> [1]
> https://lists.w3.org/Archives/Public/public-web-and-tv/2015Mar/0018.html
>
> --------
>



-- 

Yosuke Funahashi
co-Chair, W3C Web and TV IG
Web Media Specialist, W3C
Project Associate Professor, Keio University