Fraunhofer FOKUS Interest in W3C Web on TV Workshop Christian Fuhrhop Fraunhofer FOKUS Participant's interest The FOKUS Competence Center Future Applications and Media (FAME) researches and develops in the fields of Rich Media and Convergence in NGN and non-NGN IPTV environments, Human-Centric Applications for multimodal Interaction and multi-device experience, Intelligence Functions as Recommendation Systems and Communitie Services as well as Service Integration and Collaboration for mobile middleware solutions and mashup services. The main research focus is on the design and implementation of open service infrastructures. These include features as Human-Centric Applications, Service Personalization, Media Profiling, Telco Service Control and IPTV integration, Convergence, Electronic Content Guides as well as Rich Communication and interactive content approaches for IPTV. Point of View Currently there are a number of approaches to get web technology on TV, from minimalist approaches that just provide 'some sort of browser' on a set-top box to complete widget shop, storage and execution environments, or a full declarative application environment for IPTV services. And there are clearly a number of issues that are specific to TV use, such as the synchronization of applications with linear broadcasts and the possibility of broadcast distribution of widgets as part of the broadcast signal. While a certain amount of experimentation is expected during the early phases of any technology, many approaches seem to be oblivious to developments in other domains, especially the mobile domain. Additionally, current approaches seem to be based on the assumption that the current TV usage won't change significantly. It is assumed that the TV will continue to be primarily used for video viewing and Web on TV will mainly be to provide nicer teletext. Which is possibly about as true as assuming that mobile phones will be just used for phone calls and 'Web on Mobiles' is primarily a way to synchronize calendars and contact lists? And maybe look up weather and news. It is assumed that TVs are operated with a button remote control and that needs to be reflected in a Web on TV standard. But mobile phones have finger navigation, stylus navigation, virtual keyboards, numeric keypads, mini keyboards, single touch screens, multi touch screens... While not all of them are usefully employed for web browsing on mobiles, no specific input method was 'built into' the standard. It is assumed that TVs are at the low end of computing power and any significant increase of CPU, memory or graphics power would be cost prohibitive. Which might be true if all that is on offer is better teletext (where users are fairly unlikely to pay another $200 for smoother graphics), but what if someone provides a TV high-definition equivalent of World of Warcraft or Farmville? The overlap between application areas and device domains decreases. Video is watched on smart phones, the set-top box plays MP3 files, a 'new mail' reminder pops up on the phone and the user switches on the laptop to answer it. The phone builds a Skype connection to the TV. I use the EPG on my Android to program the set-top box at home. There are few 'domain specific' applications anymore, while the number of applications that either are present on all sorts of devices, perform different parts of functions of one application on different devices or allow the continuation of an application started on one device on another device is increasing . So should we really define 'domain specific' standards? And while users might not replace a TV set as often as they do replace their smart phones, the in-built capability of TV sets will become less important over time. The TV set itself may just become a HD display unit with 'Web on TV' being provided by a game console, set-top box, connected blu-ray player or dedicated entertainment computer. Suggestions We need to get out of the habit of seeing TVs as a monolithic world of its own. 'Web on TV' from a PlayStation or an Apple TV box will differ in input devices and hardware capabilities from built-in CE-HTML. Unless we want to have different fixed standards for all types of configurations, we need to define a standard that allows adaption to different kinds of input and output configurations and capabilities. And we need a standard that stays as close to the Web development in other domains as possible. For example, writing a specific HTML/XHTML version into a TV standard might simplify the situation in the TV domain in the short term, but will widen the gap to other domains in the long run. The difference between the domains becomes increasingly harder to define. If I use the Opera browser on my Wii to watch web pages on my TV, am I using 'Web on TV'? Or is that 'Web on a very slow desktop computer with no keyboard, but with fixed window size'? Should there be a standard for that? It seems more sensible to have a basic W3C standard for the Web with additional profiles, features, regulations for specific devices and situations than making a 'standards snapshot' and branching off from current developments.