The workshop was about various aspects of TV and the Web – topics included HTML5, adaptive streaming and content protection, metadata, accessibility, second screens and testing. 109 people attended, including Dan, Vicky and I from NoTube. This was the second of two W3C workshops so far on this topic – the previous one was in Japan in September last year, and there’s likely to be another one in North America. I was on the programme committee for both workshops so far and I was very pleased to be able to attend this one.
The W3C holds these workshops in order to get an idea of the potential and interest in creating working groups in this area. It’s a very interesting area because of the variety of organisations and expertise involved (e.g TV and set top box manufactuers, broadcasters, Web technologists, video player creators, browser vendors, accessibility experts, academics and TV standards experts). There are of course many existing standards organisations in TV, and many existing web standards relevant to this area.
One of the fascinating things about this area is the clash of cultures between the relatively closed world of TV and the relatively open world of the Web. Perhaps most interesting from my personal point of view is the difference in attitude towards developers. Thanks to ‘view source’ there are lots of people who develop for the web based on ‘copy and paste’ coding. I started off as one of these people, and it’s a hugely attractive aspect of the web, to be able to create content on it so easily. Many of these people rarely read specs, and if they do they expect them to be short, and trivially and freely available on the web. There are an enormous number of these poeple, many very creative in their ideas and skilled in their ability to design for users.
In contrast in the TV world, documents – and particularly standards – are not freely available and tend to be extremely lengthy and complicated. This is at least in part a product of the different market environments. People don’t update their TV hardware very often, the hardware is very low-powered and specific to its purpose, and it doesn’t get updated often because there’s no money to be made, because most purchasing models are not subscription based. Couple that with high price competition and this means that TVs and particularly set top boxes have to work very reliably on very low-powered hardware – people don’t expect their TV to just stop working. Contrast this with phones, which differ in that people tend to change them every 19-24 months allowing the manufacturers to change the functionality relatively rapidly. The requirement for reliability when devices are rarely updated or changed means that specifications must be very precise and you can’t just let any amateur developer near the hardware. These are only some of the issues – and I’m new to the TV world so only just beginning to get some sort of understanding of them – but they go some way towards explain the tendency towards closed and complex specs.
The enormous number of web-enabled TVs being sold at the moment is directly challenging these assumptions. Connected TVs have the potential for some very interesting new functionality – web on TV! – but also multiple challenges, from user-facing issues to do with the complexity of user interfaces and remotes and the effects of social watching, to multiple technical decisions and developments. For some companies and consumers, IP could be the way that much content is delivered, which is why adaptive streaming is important to some companies as it will lower their costs. As TVs and other devices get smarter there’s potential to have more standards-based web-browser technologies for display than the current profiles of aging HTML standards. There’s also the attraction of widgets and app stores, so successful on the iPhone, and potentially an important part of TV in the future.
The contrast between the two communities is clear from the position we took in our paper (pdf) for the workshop, which comes very much from a web point of view. We argued that W3C was well-placed to create guidelines around using metadata as an advertisement for programmes suitable for sharing and promotion using social media, rather than as precious metadata to be bought and sold; that it could help create guidelines for delivering resolvable http URIs with the content, again suitable for people to share and talk about; and that it could be a place to create standards around simple, developer-friendly APIs to TV. We argued that this is what the people who use these products want – as Dan put it (and got a laugh), in NoTube we’ve noticed that:
- nobody says, “I want to see recommendations from other people who bought the same TV as me”
- nobody says, “I want to learn a new interface and controller for every device I use”
- nobody says “I wish watching TV was more like using a computer.”
i.e. consumers don’t want to be trapped in incompatible silos, using complex remotes to input text on a device hat isn’t optimised for it. We also argued that there are direct benefits to content owners, developers and manufacturers.
During his presentation, Dan donated a few minutes to Matt Hammond from BBC R&D to talk about the BBC’s work on a Universal Controller API – documents about which are shortly to be made available as a white paper on the BBC’s website. The Universal Control work is conceptually very close to some of the work we have done in NoTube on the ‘Buttons’ protocol. We’ve arrived at similar needs based on somewhat different requirements – but with overlapping concerns, particularly the usablity, accessiblity, and privacy benefits of using second screens. One of the big benefits would be to open up TV control and the presentation of associated information to the kind of developer mentality that’s highly creative but perhaps not willing to read through and understand (say) 200 pages of highly technical specification.
You can see Dan’s presentation on slideshare and read our paper (pdf); minutes from the workshop are here (day 1) and here (day 2). The workshop agenda is available and the there’s also a full list of papers.