April 19th, 2002

Mobile without the Context?

This week was taken up with the Microsoft Mobility Developer Conference. A lot of technology conferences are now using the Hilton Metropole on Edgware Road, very close to both Tube stations and just off the end of the A40(M). It's a good venue, with plenty of conference and meeting space - I have to admit the occasional conrunner in me looks at it and thinks "Eastercon"! However, I'm pretty sure its room rates would work out more expensive than the Hotel De France...

The conference itself was interesting, and useful. A lot of deep technical information, though to my mind the most useful session was at the end of the conference, in which two of the leads on the Mobile Internet Toolkit platform shared their experiences and best practices for delivering cross-channel applications on the ASP.NET platform. However, I generally felt a little disappointed.

On thinking about it, it's because I felt that something was missing. I've been doing a lot of mobile development work over the last three years, and one key issue I've been trying to understand how to deal with is the idea of Context. If you've read Bruce Sterling's wonderful story "Maneki Neko" (in A Good Old Fashioned Future) you'll get a picture of a context computing driven world.

Back in my research days I spent some time working on pervasive computing concepts, and the whole ubiquitous computing model has been something of an obsession for me ever since (probably a reason for my PDA collection and my on-going love affair with the Apple Newton). It's only recently, though, that I've actually seen a way of actually fulfilling this vision. It's a bit like the moment when Charlie Stross smuggled me into the Watford SCO offices one Sunday afternoon in the early 1990s to show me Mosaic and I realised that I'd finally seen Xanadu. It's not easy yet, but a mix of rules-based personalisation and messaging, distributed component architectures, streamed serialisation, abstract UIs, and workflow are the key to delivering context-sensitive applications.

So I know there are tools and technologies that will allow us to develop applications that are sensitive to how a user wants to use them - instead of delivering the usual canned UI. And a lot of those technologies are built into the .NET platform, or can be built using it. So why did this get missed out in two days of focussed discussion and presentation. Sure, get people used to the ideas slowly, but it's a lot better to also show them just what they can do if they push the boundaries...

I want people to be excited about these things as me. And I hoped that the conference would be the place where I'd see something happen...
  • Current Music