The future of the internet revealed

The future of the internet revealed
Google TV might not be the future of television, but internet delivery almost certainly is

Technology changes so quickly, it's hard to remember how bad we used to have it. UK internet access didn't really take off until Freeserve launched in 1998, few of us had broadband before 2001 and the UK didn't even have a 3G mobile phone network until 2003.

Google was founded in 1998, Facebook and Flickr in 2004, YouTube in 2005 and Twitter in 2006. It's impossible to imagine life without them now, and the pace shows no sign of slowing down.

OnLive

STREAMING GAMES: OnLive promises to deliver console-quality gaming with the processing performed remotely

OnLive is a serious company - it boasts 200 employees, and its investors include Warner Bros and BT. It's available now in the US and looks set to grow quickly.

Cloud computing will be particularly important as smartphones and other mobile devices become the platforms of choice for most of our online activities. Phones don't yet have the power or storage necessary for desktop-calibre applications, so the emerging model is what Microsoft calls 'three screens'.

Three screens

As Steve Ballmer explains it, this is "an experience that spans the PC, the phone, the TV and the cloud". Rather than store your entire computing world on a desktop PC, you store it in the cloud and then access it on whatever device happens to be handy.

There are some things that a big desktop will almost certainly always do better than a smartphone, including data input, but there's no reason why the app has to be installed or data isolated to its own hard drive. With smartphones expected to outsell PCs by 2013 and Google's cloud-based OS Chrome on the horizon, cloud computing is going to be very important in the coming decade.

According to the Pew Internet and American Life Project, by 2020 most people can expect to "access software applications online and share and access information through the use of remote server networks, rather than depending primarily on tools and information housed on their individual, personal computers." It's all very exciting, unless you're an ISP.

Our appetite for online video is enormous and it's growing: the BBC's iPlayer delivers seven petabytes (7,000 terabytes) of video a month, while YouTube's bandwidth is estimated at 126 petabytes per month. Networking firm Cisco predicts that video will account for 90 per cent of consumer internet traffic and 64 per cent of mobile internet traffic by 2013.

Microsoft thinks online video isn't smart enough, and its solution is adaptive streaming, which it calls Smooth Streaming. Unlike traditional streaming, where your connection speed is checked once (if at all), adaptive streaming monitors your internet connection constantly.

If it becomes congested, the bitrate drops to something your connection can handle. When the congestion clears, the bitrate goes up. It works well, even on large-scale events, and you can see it in action at www.smoothhd.com.

The problem with adaptive streaming is that it still uses the old client/server model, where the server transmits data to you directly across the entire internet. BitTorrent creator Bram Cohen has an alternative idea, dubbed Project Pheon, which uses peer-to-peer networking to deliver streaming video.

Speaking at the 2010 NewTeeVee conference, Cohen promised "around five-second latency from when the content goes out to when it's actually displayed on people's machines".

Join the swarm

Pheon - like BitTorrent - uses swarming rather than traditional downloading. As you download a file, the bits you've downloaded are shared with other downloaders, so in theory you should get faster downloads by connecting to somebody near you rather than a distant server.

It's a technology that works best for popular files, and if you're a regular torrent user you'll know that new, popular torrents download like lightning while obscure ones crawl. This means swarming is best suited to big events, such as newly released films, live sports and concerts.

Of course, to actually access such high bandwidth services, we'll need fast broadband. Will we have the super-fast service the government is promising by 2017?

Trefor Davies is CTO and co-founder of business ISP Timico. "The problem facing the government is that the task is a huge one, and it would be very easy for them to decide that the only way they can realistically get to the end game is by roping in BT to help," he says, pointing out that while BT has offered to match the government's £830m funding to deliver 90 per cent super-fast broadband coverage by 2017, "coming from a company that claims to have 99 per cent broadband coverage, this makes us wonder what is meant by '90 per cent high speed broadband'."

Davies believes the only way to get fast broadband throughout the UK is to involve communities. "There are things that communities can do to make it easier and cheaper to roll out fibre networks," he explains.

"For example, companies like BT are charged anything up to £10 per metre for wayleaves to run cables across private land. [That's] a nice little earner for landowners: the average length of fibre in the Eden Valley is around 20km per community. That's a lot of wayleave charges that BT has to built into its costs." Landowners might waive those costs for community organisations, making fibre roll-out cost-effective.