Personalised TV news

Arguably (and there are often arguments), Dave Winer invented blogs. And web services. And RSS. And podcasting. There can’t have been many more influential people in the recent history of the internet. Sometimes he delivers a solution which survives; other times, he plants a seed which grows into something different. His blog, for example, doesn’t include some of the elements we’ve come to expect from ‘a blog’ – but it’s still recognisably the same notion.

So when Dave goes to the trouble of mocking up a concept for TV news of the future, it’s worth taking notice. He describes a scenario we all know too well: ‘yes, I’ve heard that news story already, I’m bored of it… can we move on, please.’ Some will want all the background, others will just want to know the top line, others won’t care in the slightest.

His visual is pretty rough, and I don’t really see it working in practice – but the mechanism he describes is certainly interesting. A big story breaks; it gets its own ‘subject tag’; and all viewers get subscribed to it automatically. When you get bored of it, you unsubscribe, and you don’t see any more about it. Presumably unless it’s a big development, in which case the cycle begins again. (Dave’s mockup also hints at some more generic subject areas: and of course he’s right, it would have to be a combination.)

It’s a while since I worked closely with online video… and I’m not sure how valid this is, now that Flash rules that territory. But in the old days, to fire off a video stream, you first had to click through a ‘playlist’ file. So to set up a RealPlayer media stream, you pointed to a .RAM file, which pointed in turn to a .RA file. I’m pretty sure Windows Media files worked in the same way, with ASX files pointing to ASF media. Few people ever used this properly; each media file had its own single-item playlist. But the potential was there (in theory at least) to generate those playlists on the fly.

This playlist concept seems to have been forgotten. I don’t even know if it’s still built into the software protocols. But it would seem that the concept is primed for a comeback.