Wednesday, 7 July 2010
This blog post can be read at http://5tth.blogspot.com
However, we should not even be thinking of the term "killer app" any more. The term came from the world of commerce who saw it as the golden goose which would lay the most profitable egg. It comes from a boom and bust, bubbly world that focuses on money before all else. It came from a need to 'justify the Internet' - how bizarre that sounds today?
From a network point of view, the killer app has also potentially had a less positive feel - could there be an app that would break the network? Some of us have always hoped that there was as it would highlight the inefficiencies and design faults within the core network and mean it would need to be fixed before ere long and hence remove that bottleneck and break the scarcity model once and for all!
Killer App becomes Killer Cap, as in capacity, capability?
I digress...I have been thinking about convergence and divergence recently. Convergence because we are all seeing new devices etc which overlap and bring unexpected features and functionality into easy reach. Divergence because we are moving away from expected models and futures because of the convergence.
For those of us who have been banging on about Fibre To The Home, fat pipes, and FiWi (increasingly), there has never been a need for a single killer app to prove our point that people need to be able to do precisely what they want to do over a network wherever they are, whenever they choose to and however they want to do it.
People will do what they choose and applications will rise and fall in popularity, meaning that the network must be sufficiently capable of dealing with the vagaries of human nature and differing interests far more than it needs to be able to cope with a killer app. (It also needs to cope with "consumers are also creators", and asymmetry is the biggest problem here once people have sufficient bandwidth and the devices to share what they have created.)
However, convergence has meant that the mobile phone has taken a starring role in driving forward application development and hence network usage. Mobiles have introduced more people to technology than computers have. Whilst digital video cameras may be expensive, the video on your mobile is usually more than sufficient to 'capture a moment' and MMS, Bluetooth etc have meant sharing it, whether via the Internet eg Youtube, Facebook etc or directly to another handset, is something that most people can master reasonably easily.
And herein lies the divergence, and, to me, the importance of FiWi. FiWi is Fibre ---> Wireless. (Many people forget that a mobile phone is a wireless device but for the purpose of this article, wireless also encompasses all the different uses of the wireless spectrum - wi-fi, Wimax (RIP?), GSM etc.)
A fat pipe is required for backhaul and the core network. There are no two ways about that, especially as our data transport requirements soar by the hour. If you cannot send enough packets through the network, consumers will moan. And rightly so. However, where the fat pipe reaches to only becomes an issue if the device and application(s) at the end of the first mile e.g. under the consumer's fingertips are using bloated code.
Mobile applications have necessarily used tight programming and compression techniques. Firstly, because of the internal memory limitations of the device, but also because the mobile network has limited capacity. The spectrum is a finite resource. You can't be letting people send 20 photos of granny's birthday party which are all 50MB, f'r instance.
But web applications have been developed with increasing amounts of memory eg in personal computers and servers, and increasing amounts of bandwidth available to the consumer to send or receive data. This has allowed web programmers to become 'lazy'. Why develop awesome compression techniques for videos if the connectivity a typical consumer has can handle what is already existent, and the likelihood is that this connectivity is only going to get fatter?
WAP was a much-maligned protocol but actually it was thinking along the right lines. People are going to use wireless to access the Net. More so now than ever before as the convergence becomes inherent in our thinking. iPads, iPods, smartphones and even your average £10 mobile handset can all access the Net. The idea that to do so would require you to plug in to any type of socket is now anathema to all of us.
Whether the wireless is in the first inch ie the wireless access point that spews wireless around your house or in the Bluetooth you use to share your photos, or the first mile ie to overcome technical difficulties with laying a fixed line solution, or even a satellite connection (which is also wireless), if you then connect this to fibre, you have the answer to the problem.
What problem? Oh, well, that'll be how to get fibre as close to every single human as possible.