- From: Bob Lund <B.Lund@CableLabs.com>
- Date: Mon, 11 Apr 2011 14:25:17 -0600
- To: Matt Hammond <matt.hammond@rd.bbc.co.uk>, Clarke Stevens <C.Stevens@CableLabs.com>, "public-web-and-tv@w3.org" <public-web-and-tv@w3.org>, Olivier Carmona <ocarmona@awox.com>, Russell Berkoff <r.berkoff@sisa.samsung.com>
- CC: Giuseppe Pascale <giuseppep@opera.com>
Hi Matt, See inline below. Regards, Bob > -----Original Message----- > From: Matt Hammond [mailto:matt.hammond@rd.bbc.co.uk] > Sent: Monday, April 11, 2011 9:49 AM > To: Clarke Stevens; public-web-and-tv@w3.org; Olivier Carmona; Russell > Berkoff; Bob Lund > Cc: Giuseppe Pascale > Subject: Re: [HOME_NETWORK_TF] Home Network Technologies > > Hi Bob, > > Yes that is exactly what I had in mind - the ability query and control > those aspects of the TV experience - not just those involving streaming > of media from other devices in the home. For the BBC it would be > important to support the use cases I outlined for broadcast content that > potentially never leaves the receiver device. > > Did anything come of Doug's suggestions? What kind of model did he have > in mind? The concept of "remote control" could have several meanings: > > * the simulation of button pushes or other means of driving the existing > UI presented on the screen by the TV itself > It is this model - how a remote control device interacts in a generic manner with a device presenting HTML based UI. How the remote and UI device "discover" each other is out-of-scope. A common "discovery" scenario is I point my remote at the UI device I want to control. > * a content centric model for querying and control in the style of UPnP > services, Universal Control API, or others > > The latter better lends itself to supporting dual-screen scenarios. > Forgive me if I'm restating what is already obvious but the UPnP services allow a UI device to discover media players and servers, each of which are separate devices. A single UI device can control a multiplicity of servers connected to players. Thus, the UI device is in many senses a remote control. This UI device can send user input to a server or player that itself is presenting UI. So there is support for two paradigms: a remote controlling a UI device that is probably right in front of you and a UI device controlling multiple servers and renders anywhere on the home network. > > regards > > > > Matt > > On Fri, 08 Apr 2011 20:25:46 +0100, Bob Lund <B.Lund@cablelabs.com> > wrote: > > > Hi Matt, > > > > So if I understand correctly, you're thinking about a service to > > control the TV - content never leaves it. This UPnP service > > http://upnp.org/specs/av/UPnP-av-RenderingControl-v2-Service.pdf > > provides a remote interface to some controls of a rendering device but > > probably not all of the controls a TV with a tuner would have. > > > > Sometime ago when Web and TV had its first workshop, Doug Schepers > > solicited input for remote control events related to TV with the > > intent, I believe, of defining DOM3 events for these. This would > > enable remotes to work across UI implementations running on compliant > user agents. > > > > Bob > > > >> -----Original Message----- > >> From: Matt Hammond [mailto:matt.hammond@rd.bbc.co.uk] > >> Sent: Friday, April 08, 2011 9:59 AM > >> To: Clarke Stevens; public-web-and-tv@w3.org; Olivier Carmona; > >> Russell Berkoff; Bob Lund > >> Cc: Giuseppe Pascale > >> Subject: Re: [HOME_NETWORK_TF] Home Network Technologies > >> > >> Hi Bob, > >> > >> Your latter example is what I had in mind: a device that is connected > >> to the home network and receives off-air broadcasts, but does not > >> serve or stream that content around the home network. Instead the > >> device itself includes a display, or is connected directly to one > >> using HDMI or similar. > >> > >> It is not clear to me if the A/V transport service (or whichever > >> service is appropriate) facilitates remote control type functions > >> such as changing channel or playing back a recording, as distinct > >> from choosing content to stream to another device. > >> > >> regards > >> > >> > >> > >> Matt > >> > >> On Fri, 08 Apr 2011 15:30:58 +0100, Bob Lund <B.Lund@cablelabs.com> > >> wrote: > >> > >> > Matt, > >> > > >> > See inline below. > >> > > >> > Bob > >> > > >> >> -----Original Message----- > >> >> From: Matt Hammond [mailto:matt.hammond@rd.bbc.co.uk] > >> >> Sent: Friday, April 08, 2011 3:03 AM > >> >> To: Clarke Stevens; public-web-and-tv@w3.org; Olivier Carmona; > >> >> Russell Berkoff; Bob Lund > >> >> Cc: Giuseppe Pascale > >> >> Subject: Re: [HOME_NETWORK_TF] Home Network Technologies > >> >> > >> >> We also see facilitating discovery of content on the home network > >> >> from the browser to be important. There is also the issue of > >> >> security and privacy - a user should be able to determine what > >> >> devices, applications and web pages have permission to control > >> >> another device and discover content. > >> >> > >> >> As a broadcaster we deliver to multiple platforms over which we > >> >> have no direct control (terrestrial, cable and satellite > >> >> transmission all operated by different companies/consortia). Our > >> >> interest is closer to the scenario > >> >> 2(b) that you outline. > >> >> > >> >> What is not clear to me is whether the DLNA/UPnP capabilities for > >> >> discovering and playing content also work for a box that does not > >> >> have the capability to serve or stream off-air broadcast content > >> >> on the home network (for example, a TV with integrated tuner). > >> > > >> > I am not clear on the question. If you mean a device that is not > >> > connected to the home network then UPnP does not address that. But, > >> > such a device would seem to be out of scope of all home network > APIs. > >> > If you mean a device that is connected to the home network but does > >> > not have the capability to serve or stream off-air broadcast > >> > content, then yes, the device can host the content discovery, > >> > connection manager and A/V transport services. A TV with an > >> > integrated tuner could serve off-air broadcast content if it was > >> > also connected to the home network and had the appropriate UPnP > services. > >> > > >> > > >> >> This is one of the > >> >> reasons why we have an interest in a simplified abstraction that > >> >> negates the need for a controlling device to concern itself with > >> >> the source and delivery method of a piece of content (as well as > >> >> its format, codecs, device profiles etc) > >> >> > >> >> There is also potential to enable a substantial accessibility win > >> >> here too if UI can be independent of the device(s) it is > controlling. > >> >> > >> >> When considering the Remote UI as mentioned here, perhaps there > >> >> are at least 3 related strands to this? Discovery/security; remote > >> >> control; device-served remote UI. > >> >> > >> >> regards > >> >> > >> >> > >> >> > >> >> Matt > >> >> > >> >> On Thu, 07 Apr 2011 20:14:19 +0100, Bob Lund > >> >> <B.Lund@cablelabs.com> > >> >> wrote: > >> >> > >> >> > To amplify on Clarke's point a bit, it's useful to break down > >> >> > the broadcast/cloud UI into more specific scenarios: > >> >> > 1) Broadcast content directly from the cloud, UI from the cloud > >> >> > 2) Broadcast to a UPnP media server that proxies the content on > >> >> > the home network via content directory service, remote UI from > >> >> > the cloud > >> >> > > >> >> > Scenario 1 is the web, nothing special needs to be introduced. > >> >> > Scenario 2 is more of a DLNA/UPnP model. There are two ways the > >> >> > UI served from the cloud can gain the links to refer to the > >> >> > media server > >> >> > content: > >> >> > a) A back channel from the media server hosting the CDS to the > >> >> > cloud server constructing the UI page. If the service provider > >> >> > controls both the media server software implementation and the > >> >> > cloud UI server this back channel can be service provider > specific. > >> >> > The cloud UI server constructs UI that references the home > >> >> > network content. There are some subtleties around how the media > >> >> > LAN side IP address is referenced in the UI page but it's > doable. > >> >> > b) The UI page served by the cloud server to the home network > >> >> > device contains JavaScript that uses the discovery API that is > >> >> > being proposed to discover the CDS and the content therein. The > >> >> > UI page JavaScript can provide whatever degree of integration is > >> >> > desired between the local content discovered and the cloud > >> >> > content referenced on the UI > >> >> page. > >> >> > > >> >> > Bob Lund > >> >> > > >> >> >> -----Original Message----- > >> >> >> From: public-web-and-tv-request@w3.org > >> >> >> [mailto:public-web-and-tv- request@w3.org] On Behalf Of Clarke > >> >> >> Stevens > >> >> >> Sent: Thursday, April 07, 2011 11:05 AM > >> >> >> To: Matt Hammond; public-web-and-tv@w3.org; Olivier Carmona; > >> >> >> Russell Berkoff > >> >> >> Cc: Giuseppe Pascale > >> >> >> Subject: RE: [HOME_NETWORK_TF] Home Network Technologies > >> >> >> > >> >> >> Matt, > >> >> >> > >> >> >> UPnP currently defines a way to discover remote user > >> >> >> interfaces, but it is pretty generic. Here are the specs: > >> >> >> > >> >> >> http://upnp.org/specs/rui/remoteui/ > >> >> >> > >> >> >> In this group, we're hoping to make this a bit more concrete by > >> >> >> using HTML browsers, CSS, JavaScript, and providing some APIs > >> >> >> for discovering and using devices and services on the home > network. > >> >> >> The back channel to the cloud service would just be a URL. > >> >> >> That's maybe a bit oversimplified as there are other issued > >> >> >> related to a commercial service offering, but that's the basic > idea. > >> >> >> > >> >> >> As for more information on DLNA CVP, here's an announcement: > >> >> >> > >> >> >> http://www.dlna.org/news/pr/view?item_key=e2c163bfab8076edc2b33 > >> >> >> eba > >> >> >> 829 > >> >> >> 3e8 > >> >> >> 2cd2f11e3e > >> >> >> > >> >> >> It looks like DLNA doesn't yet have CVP for sale from its web > >> >> >> site but W3C could encourage them to provide some more > >> >> >> information, timelines, etc. > >> >> >> > >> >> >> Hope this helps. > >> >> >> > >> >> >> Thanks, > >> >> >> -Clarke > >> >> >> > >> >> >> -----Original Message----- > >> >> >> From: Matt Hammond [mailto:matt.hammond@rd.bbc.co.uk] > >> >> >> Sent: Thursday, April 07, 2011 8:11 AM > >> >> >> To: public-web-and-tv@w3.org; Olivier Carmona; Russell Berkoff; > >> >> >> Clarke Stevens > >> >> >> Cc: Giuseppe Pascale > >> >> >> Subject: Re: [HOME_NETWORK_TF] Home Network Technologies > >> >> >> > >> >> >> Hi Clarke, > >> >> >> > >> >> >> > >> >> >> Could you explain how this model where a remote user interface > >> >> >> is served from the cloud would work for broadcast services? > >> >> >> Does it require some kind of back channel between the rendering > >> >> >> device and > >> >> the cloud service? > >> >> >> > >> >> >> I would also like to understand what services CVP enables. Are > >> >> >> the guidelines publicly available anywhere? > >> >> >> > >> >> >> regards > >> >> >> > >> >> >> > >> >> >> > >> >> >> Matt > >> >> >> > >> >> >> > >> >> >> On Thu, 07 Apr 2011 13:40:12 +0100, Clarke Stevens > >> >> >> <C.Stevens@cablelabs.com> wrote: > >> >> >> > >> >> >> > I think some of the functionality you want is already defined > >> >> >> > in DLNA/UPnP and the rest will be very soon. As I mentioned > >> >> >> > earlier, the URL that points to the content does not need to > >> >> >> > be on the home > >> >> >> network. > >> >> >> > This allows for content that can be offered from service > >> providers. > >> >> >> > Similarly, a service provider could serve a remote user > >> >> >> > interface from the cloud. Features such as volume change and > >> >> >> > mute are already provided, so a "second screen" could get a > >> >> >> > user interface from a service provider and point to content > >> >> >> > remotely served by that service > >> >> >> provider. > >> >> >> > > >> >> >> > > >> >> >> > There is also a DLNA guideline called CVP specifically > >> >> >> > designed for content service providers. The first version is > >> >> >> > available now but is somewhat limited. The second version is > >> >> >> > in development and I believe it provides the remainder of the > >> >> >> > services you are looking > >> >> for. > >> >> >> > > >> >> >> > Thanks, > >> >> >> > -Clarke > >> >> >> > > >> >> >> > -----Original Message----- > >> >> >> > From: public-web-and-tv-request@w3.org > >> >> >> > [mailto:public-web-and-tv-request@w3.org] On Behalf Of Matt > >> >> >> > Hammond > >> >> >> > Sent: Thursday, April 07, 2011 2:52 AM > >> >> >> > To: public-web-and-tv@w3.org; Olivier Carmona; Russell > >> >> >> > Berkoff > >> >> >> > Cc: Giuseppe Pascale > >> >> >> > Subject: Re: [HOME_NETWORK_TF] Home Network Technologies > >> >> >> > > >> >> >> > Hi Russell, > >> >> >> > > >> >> >> > Your description matches my understanding of DLNA/UPnP - > >> >> >> > which is useful confirmation for me - many thanks! > >> >> >> > > >> >> >> > I agree that this model works well for streaming between home > >> >> devices. > >> >> >> > But it makes the assumption that all content is streamable > >> >> >> > between home devices. A TV or set-top-box that can receive > >> >> >> > broadcast or IP delivered TV or radio does not necessarily > >> >> >> > match this. In order to be able to implement some of the > >> >> >> > second-screen scenarios presented at the last workshop, I > >> >> >> > would hope that APIs for home networking are also able to > >> >> >> > query and control TV functions such as changing channel or > >> >> >> > retrieving broadcast > >> programme metadata. > >> >> >> > > >> >> >> > This is why I am interested in APIs that can support both in > >> >> >> > (if > >> >> >> > possible) a single simple unified way. The approach we took > >> >> >> > at the BBC with our Universal Control (UC) API is one > >> >> >> > possible way to achieve > >> >> >> this. > >> >> >> > > >> >> >> > If a TV implementing UC API needed to be able to discover and > >> >> >> > stream content on other devices in the home then it would > >> >> >> > almost certainly use DLNA/UPnP or other similar technologies > >> >> >> > to do this, but the client/controller device does not need to > >> >> >> > know the difference. From the perspective of a client of the > >> >> >> > UC API implementing TV, streamed content is discoverable and > >> >> >> > selectable for playback in the same way as a television > >> >> >> > channel or programme > >> >> would be. > >> >> >> > > >> >> >> > regards > >> >> >> > > >> >> >> > > >> >> >> > > >> >> >> > Matt > >> >> >> > > >> >> >> > On Thu, 07 Apr 2011 08:39:18 +0100, Russell Berkoff > >> >> >> > <r.berkoff@sisa.samsung.com> wrote: > >> >> >> > > >> >> >> >> Hi Matt, > >> >> >> >> > >> >> >> >> > >> >> >> >> Let me try to answer "frame" your question/observation and > >> >> >> >> hopefully provide some brief answers: > >> >> >> >> > >> >> >> >> > >> >> >> >> Q1: Does DLNA/UPnP distinguish between the description of > >> >> >> >> content, > >> >> >> i.e. > >> >> >> >> metadata describing content (title, description, available > >> >> >> >> media > >> >> >> >> formats) and the device actually supplying the content? > >> >> >> >> > >> >> >> >> > >> >> >> >> Q2: Does DLNA/UPnP allow metadata and content to reside on > >> >> >> >> separate devices on the home network? > >> >> >> >> > >> >> >> >> > >> >> >> >> Q3: Does DLNA/UPnP allow the remote commanding of playback > >> >> devices? > >> >> >> >> > >> >> >> >> > >> >> >> >> > >> >> >> >> A1: Yes...DLNA/UPnP has a distinct metadata service > >> >> >> >> (ContentDirectory Service). The metadata provided by this > >> >> >> >> service can refer to content elsewhere on the home network. > >> >> >> >> > >> >> >> >> > >> >> >> >> A2: Yes > >> >> >> >> > >> >> >> >> > >> >> >> >> A3: Yes...UPnP/DLNA allows remote commanding of playback > >> devices. > >> >> >> >> A Digital Media Controller can obtain descriptive content > >> >> >> >> metadata and tell a playback device to "pull" a piece of > >> >> >> >> content from a distinct "content source" on the home > network. > >> >> >> >> > >> >> >> >> > >> >> >> >> The Digital Media Controller in this scenario would examine > >> >> >> >> the metadata from both the Playback Device (Digital Media > >> >> >> >> Renderer) and from the metadata service (Digital Media > >> >> >> >> Server - ContentDirectory > >> >> >> >> Service) and insure a media format is chosen that both the > >> >> >> >> Content Source can provide and the Playback Device can > render. > >> >> >> >> Once a match is found the Digital Media Controller, would > >> >> >> >> command the Playback device to fetch the content from the > >> Content Source. > >> >> >> >> > >> >> >> >> > >> >> >> >> Regards, > >> >> >> >> > >> >> >> >> Russell Berkoff > >> >> >> >> > >> >> >> >> Samsung Electronics > >> >> >> >> > >> >> >> >> > >> >> >> >> > >> >> >> >> > >> >> >> >> -----Original Message----- > >> >> >> >> From: Matt Hammond [mailto:matt.hammond@rd.bbc.co.uk] > >> >> >> >> Sent: Wednesday, April 06, 2011 5:03 AM > >> >> >> >> To: public-web-and-tv@w3.org; Olivier Carmona > >> >> >> >> Cc: Russell Berkoff; Giuseppe Pascale > >> >> >> >> Subject: Re: [HOME_NETWORK_TF] Home Network Technologies > >> >> >> >> > >> >> >> >> > >> >> >> >> Hi Olivier, > >> >> >> >> > >> >> >> >> > >> >> >> >> On Wed, 06 Apr 2011 12:22:38 +0100, Olivier Carmona > >> >> >> >> <ocarmona@awox.com> > >> >> >> >> > >> >> >> >> wrote: > >> >> >> >> > >> >> >> >> > >> >> >> >>> > >> >> >> >> > >> >> >> >>> Hi Matt, > >> >> >> >> > >> >> >> >>> > >> >> >> >> > >> >> >> >>> DLNA 1.0 is about two-box pull model: on one side you have > >> >> >> >>> a Digital > >> >> >> >> > >> >> >> >>> Media Player (a client in your description) and on the > >> >> >> >>> other side you > >> >> >> >> > >> >> >> >>> have a Digital Media Server (a rendering device in your > >> >> >> description). > >> >> >> >> > >> >> >> >>> DMP discovers and then browses DMS, and can request one of > >> >> >> >>> the browsed > >> >> >> >> > >> >> >> >>> items to be played. The only difference with your model is > >> >> >> >>> that this the > >> >> >> >> > >> >> >> >>> DMP that based on the information exposed by the DMS, > >> >> >> >>> decides wherever > >> >> >> >> > >> >> >> >>> it can play the content. > >> >> >> >> > >> >> >> >> > >> >> >> >> My explanation was not as clear as it should have been - > >> >> >> >> please accept my > >> >> >> >> > >> >> >> >> apologies for that. The client that I described does not > >> >> >> >> display media and > >> >> >> >> > >> >> >> >> does not send or receive media streams. It is the server > >> >> >> >> that will > >> >> >> >> > >> >> >> >> render/display media, and it is up to the server to work out > >> >> >> >> what media is > >> >> >> >> > >> >> >> >> available to it and to arrange to stream (or otherwise > >> >> >> >> obtain) > >> it. > >> >> >> >> This > >> >> >> >> > >> >> >> >> aspect would be completely opaque to this client. The > >> >> >> >> server/renderer > >> >> >> >> > >> >> >> >> could well use DLNA to discover and stream that content from > >> >> >> >> a 3rd device, > >> >> >> >> > >> >> >> >> but the client would not be aware of this. > >> >> >> >> > >> >> >> >> > >> >> >> >> For me, the attraction of this model is that it is more > >> >> >> >> abstract that just > >> >> >> >> > >> >> >> >> streaming and can easily subsume access to local storage but > >> >> >> >> also access > >> >> >> >> > >> >> >> >> to services such as live television broadcast. A television > >> >> >> >> programme is > >> >> >> >> > >> >> >> >> just another item of content that the server reports it has > >> >> >> >> available to > >> >> >> >> > >> >> >> >> it, and which the client can ask it to display. > >> >> >> >> > >> >> >> >> > >> >> >> >> > >> >> >> >> > >> >> >> >> regards > >> >> >> >> > >> >> >> >> > >> >> >> >> > >> >> >> >> Matt > >> >> >> >> > >> >> >> > > >> >> >> > > >> >> >> > >> >> >> > >> >> >> -- > >> >> >> | Matt Hammond > >> >> >> | Research Engineer, BBC R&D, Centre House, London > >> >> >> | http://www.bbc.co.uk/rd/ > >> >> > >> >> > >> >> -- > >> >> | Matt Hammond > >> >> | Research Engineer, BBC R&D, Centre House, London > >> >> | http://www.bbc.co.uk/rd/ > >> > >> > >> -- > >> | Matt Hammond > >> | Research Engineer, BBC R&D, Centre House, London > >> | http://www.bbc.co.uk/rd/ > > > -- > | Matt Hammond > | Research Engineer, BBC R&D, Centre House, London > | http://www.bbc.co.uk/rd/
Received on Monday, 11 April 2011 20:26:15 UTC