Saturday, November 22, 2008

BodyNet fifteen years after

About fifteen years ago, Olin Shivers looked at a typical tech-savvy professional of the day carrying a pager (remember those?) a cell phone, a "digital diary", a keyless car remote, a notebook computer and a Walkman (remember those?) and concluded "That's one headset, two communications systems, four keyboards and five displays too many." You typically didn't have a good mobile web connection in those days, either. Shivers then went on to describe a more modular collection of pieces communicating through a short-range network he dubbed "BodyNet".

Fast-forward fifteen years. Are things any better? Well, yes. Did BodyNet happen? Well, depends on how you count.

These days you can get a portable thingie [see also "pocket-thing"] that will let you make phone calls, download and play music and videos, get and answer email, surf the web, keep your calendar and use GPS (or other means) to tell you where you are. You can also download other widgets/gadgets/apps/whatever-you-call-them that will let you do all manner of other things (or at least, play games).

There is even a short-range network standard, namely Bluetooth, that you can use to attach accessories to the thingie, including a headset that will let you talk hands-free, albeit at the cost of sometimes looking like you're having a conversation with the wall or your invisible friend Harvey the Rabbit. Except for its somewhat broader range, Bluetooth looks remarkably like the BodyTalk Shivers describes. I doubt that's a coincidence.

So: The hodge-podge of mobile devices one carries around have consolidated. There is now a short-range, personal, body-sized network. This network can connect to the Web. There is a market for devices to plug into that network. So BodyNet happened, right?

Not really. The original BodyNet was meant to be a mix-and-match affair in which you get the pieces you need a la carte and plug them together. Shivers specifically argues against a "monolithic" approach:
We do not believe in this [monolithic approach] over a broad class of users [...] Individual users will persist in remaining individual -- the system requirements of one will not satisfy the needs of another.
But the monolithic approach is just what we have today. "Phones" pack in more and more features. Worse (from the geekly perspective), they tend to do so in a very closed-system sort of way. The cell phone makers and service providers want very tight control over what you can and can't do with their product/service. True mix-and-match is restricted to accessories like headsets.

This seems to be a recurring blind spot for us geeks. We're taught "clean interfaces", "standard protocols" and "modularity, modularity, modularity." We forget that most people don't want to pick their favorite components and plug them together. The best selling stereos (do people still listen to stereos?) are all-one-piece. We have software "office suites" because no one wants to find out how well Word Processor A works with Spreadsheet B. Even on the wild and woolly web, people like portals and mashups that do the grunt work of pulling things together.

Phone makers sell all-singing, all-dancing phone/modem/GPS/email/web/music/video/... devices because people want them. They want the features, not the joy of putting them together.

No comments: