Rupert Goodwins' Diary

Monday 10/10/2005
Up in Cambridge, where I was taking part in a Cambridge MIT Institute day of wireless future roadmapping. I was there as much to throw ideas in as to take notes and a lot of what was said was off the record, so some things will have to remain secret.

One such thing is the identity of the chap who pointed out that Google's policy of continuously increasing Gmail storage allocations works out to around 1.6 kilobits/second per mailbox -- so by aggregating a couple of hundred free accounts, you can create a data store capable of absorbing a video stream. Indefinitely. None of the Gmail filing systems I know can aggregate multiple accounts, but it's only a matter of time -- is this the backup problem finally fixed?

There then followed a discussion about the BBC's new Internet broadcasting experiment, iMP. This is a cunning wheeze that tries to get around the problem of huge peak bandwidth demands for streamed video by using peer-to-peer. Streamed P2P is not a goer, because the random delays in moving stuff into then out of a peer soon mess things up: instead, the BBC is using DRM to push encrypted programmes out to people ahead of time. If you say you're always going to want to watch East Enders, for example, then you can get the programme when it's ready days ahead of schedule, delivered by a torrent over a few hours. All the BBC has to do at broadcast time is send you a few hundred bytes of key.

Other tricks up Auntie's lace-trimmed sleeve include variable encoding depending on the popularity and content of the programme -- Checkout Challenge will be compressed more ruthlessly than Brian Sewell's Top Ten Tintorettos -- and the magical DRM-powered disappearance of content after seven days. That's fair enough, but do they have to use Microsoft's DRM? No they don't but yes they are, at least for the trial.

This was all over coffee -- we then got down to the serious business of working out what was going to happen next and whether this was a good idea or not. The odd thing about wireless industry luminaries is that they are a bit, well, odd. At one stage, we were divided into groups of four to concentrate on particular aspects of the technology: the group that got something to do with fast data services to mobile phones discovered that two of them didn't own or use a mobile phone anyway. As for my group -- we swiftly worked out that 75 percent of us were radio amateurs: a fatal realisation as far as the remaining 25 percent went.

I can exclusively reveal one breakthrough fact: the port at Queens' College is really very fine and has peculiar space-time dilatory effects. It makes the Holloway Road seem very far away indeed.