IPTV/VoD: The open fourth platform
Rome wasn't built in a day...
Traditional platforms take months of training and testing to build anything the brand owners will accept, and cost a small fortune to even get involved in. Sky noticed it, hence the "Sky Net" service and micropayment mechanisms.
Real-time DVD-quality video from within in a web application is a developer's wet dream – one that the likes of YouTube and Google Video have been edging us closer to for some time now. Video ondemand is simply digital video files (AVI, MPG, MOV etc) being streaming over an QoS-enabled IP network using RTP/RTSP, and live video is just another of those streams, with its source set to multicast IP address and controlled with the likes of IGMP.
With this kind of technology already freely available to anyone who looks for it, we have the basis of an open TV platform that anyone can develop for – even the weekend hobbyist. Access to that platform needs to be wide open so it is truly available equally to all. Anyone's IPTV service should be accessible on any IPTV platform by any subscriber anywhere.
Indeed, one of the first pioneers of IPTV services in the UK are the academic community. Dozens of universities and colleges already provide high-bandwidth network connectivity to students across multiple campuses that they typically use for trading terabits of illegal music and movies (with Direct Connect, or DC++), when not finishing their assignments.
Small ISPs that are involved in specialised, targeted local-loop unbundling are plugging terrestrial TV aerials into their networks and serving up multicast TV and radio over wide area IP networks onto student PCs and into communal living areas.
It's a booming business, as the same systems can be extended to hotels and new property developments. Proprietary system vendors are running their hands in anticipation of getting their hands on the notfor-profit cheque book.
The first choice IT managers at universities face is whether to build a TV system from their beloved open-source products or buy something in that somebody has already made. What makes these projects interesting is that the heavy-duty localised infrastructure is already in place and also that the motivation for building them is not typically commercial and so you tend to have more flexibility and possibilities for innovation.
The usual requirement is for putting Freeview over the LAN as IP multicast and propagating the TV channel their media students run as part of their course studies. What they want to do is usually possible with over-the-counter PC components available on the high street (such as TV tuner cards) and free software (VLC etc).
What these organisations need to do is collaborate to build the core of a standardised, open TV platform themselves, based on open source technology that costs very little and allows students to innovate around it to their hearts content. Building an open structure (e.g. XML API for scheduling and data tools, listing of multicast addresses, bandwidth allowances and traffic shaping, access to network drives, providing code for applications etc) and allowing students to play obviously provides academic benefit (or "value" as moronic businesspeople would say). Working together also helps prevent unnecessary duplication of resources and expenditure.
It's easy to see why ISPs are interested in building their IPTV networks, as they have the skills inhouse and are proficient when it comes to network systems. Theoretically speaking, we should be able to see over 200 new individual TV platforms of different sizes built and launched that you could buy over the counter in Dixons, just as you can Sky Digital or Wanadoo broadband.
There is no reason why you should have to use more than one set-top box, in the same way that you don’t have to swap mobiles when you change mobile provider. Boxes should be able to be "unlocked" and pre-installed with a giveaway CD from a shop. Content issues aside (and they are not small issues at all), all would be built on similar systems, using the same languages, messaging, back-end servers and infrastructure principles.