The Shape of Days

I was going to write a big, long post tonight about Android, the new whatever-the-fuck-it-is that Google announced today to try to grab some microscopic fraction of the attention Apple’s generated with the iPhone.

But a friend of mine let me know this morning that I’ve been writing about technology too much lately.

That’s okay. The gist of the post was going to be this: Android isn’t anything. Literally. It’s not a product, it’s not a service, it’s not a platform. At best — and I’m being uncharacteristically generous here — it’s a sort of vague specification for what basic features a phone might someday have, with some not-even-partly finished thoughts about how those features might be glued together with software.

Except being generous like that, and giving Google the benefit of the doubt, gets us into some trouble. Because to the extent that Android is anything — not so much a thing in and of itself, but a suggestion of potential things someday — it sucks.

There are two basic problems with Android. The first is that it’s based on two utterly obsolete pieces of technology: Linux and Java. If you don’t know what those are, you’re in good company. They’re also basically nothing.

At one time, Linux was a sort of hobby kit for computer enthusiasts, kind of a latter-day equivalent of the erector sets that postwar nerds thought were hot shit. Then the hype machine exploded around the turn of the century, and a lot of otherwise very smart people bought into the notion that Linux might someday be something more. What exactly was unclear, but the pie-in-the-sky, venture-capital-consuming business plans were knee-deep from Milbrae to San Jose. None of that ever played out, primarily because Linux isn’t really very good at being anything other than a least-common-denominator for sticking microchips together, but also because some political activists and lawyers of questionable ethical standing got involved and tried to boss everybody around. Despite the hostile rhetoric and the lawsuits and threatened lawsuits, Linux managed to carve itself out a sort of niche in embedded systems — you know, the little computers that run things like your cable box. If you’ve ever tried to use a cable box these days, you already have a deep and intuitive grasp of just what a mistake Google’s made here.

The other thing, Java, is also probably something you’ve never heard of. At one time — and I’m talking like twelve years ago here, ancient history in technological terms — Java was proposed as a sort of hardware- and platform-independent runtime for user applications. Except it could only be used to create really terrible user applications. Java also managed to carve a couple of niches out for itself, although one of them is closing fast. Java today is used in precisely two contexts: For writing entirely automatic computer programs to hook big databases together, like for doing airline reservations and banking and stuff, and for writing useless little crap programs that run on useless little crap phones.

If Google’s staffers had gone out of their way to pick two pieces of foundational technology that would make the Android project seem more dated and out-of-touch, they would have been hard pressed. In order for the underpinnings of Android to be any less relevant, the platform would have to be based on COBOL and … like … steam or something. Carrier pigeons, maybe. Or smoke signals.

Anyway, that’s one fundamental mistake that Google has apparently made, and not only made, but made very publicly: They’ve decided to build this supposedly shit-hot new thing on a foundation that’s rickety, decrepit and that never really worked very well to begin with.

But there’s another huge mistake here, and I haven’t yet seen very many people articulate it. I’m sure that’s because it’s just so blindingly obvious that, to everybody with an opinion on this subject, it just goes without saying: Google is going about this all backwards.

Here’s how you build a really good consumer electronics product: Decide what the user experience needs to be like, figure out what capabilities it needs to have to deliver that user experience, then fit various pieces of technology together to deliver those capabilities. When compromises have to be made — and they will — always sacrifice technology first, capability second and user experience last.

Here’s what Google’s doing with their Android thingamajig: Create a laundry list of technology components that sound cool, spackle them together with as little software as possible and as much hope as you can scrounge up, and let the user experience fall where it may.

Google’s wish-list for a common mobile phone platform reads like it was written by an electrical engineer, which it almost certainly was: cameras of such-and-such specification that can do stills and video, wireless of some variety, 3G, GPS, Java, blah blah blah. Conspicuously absent from any of that is a vision for what all those gadgets will be good for, or whether the experience of having them all glommed together like that will be pleasant or unpleasant. Here’s just one example: If you smoosh off-the-shelf components for wireless Internet and 3G radio together with a microprocessor that’s fast enough to run extremely sluggish Java applications in a way that anybody will like, you’re going to end up with a phone that’s small and has a two-hour battery, or one that can run all day but weighs eleven ounces. By starting with the technical specification — the parts manifest, essentially — and ignoring the user experience, Google is recommending that phone makers sacrifice customer satisfaction for marketing buzzwords. Which, I suppose, is what the phone makers have been doing all along, so that’s really no big shocker.

The biggest mistake Google has made here — and I’m just going by the company’s various public statements on the subject — is assuming that people want mobile phones that are “open.” They don’t.

What’s that you say? You really do want a phone that’s “open?” Then answer this: I give you a choice between two phones. One is reliable and dependable with a reasonable battery life and acceptably stable software, but it’s “closed.” That is to say, you can’t screw around with the software on it. The other phone is “open,” and you’re welcome to screw around with the software, but it doesn’t work very well. It drops calls, it gets lousy reception, it’s hard to use, the battery life sucks. Those are, for the purposes of our thought experiment, the only two phones in the world.

Which do you choose?

You choose the phone that works, obviously. Because when it comes right down to it, a phone isn’t a toy. It’s a utility device. It has to, above all else, dependably perform the primary function for which you purchased it. Maybe that primary function was to make and receive calls; maybe it’s text messaging; maybe it’s e-mail. Whatever. If the phone doesn’t dependably perform the primary function for which you bought it, it sucks. Period.

Everything else is gravy.

Customers don’t want phones that are “open,” because being “open” isn’t anybody’s primary function. If you care about it at all, it’s pretty far down on your list compared to things like “can make calls reliably” and “battery lasts all day” and “doesn’t weigh eight pounds.”

But Google has put “open” as item one on their list of features for Android. Which means everything else is secondary. If Google could make a better mobile-phone operating system that’s not “open,” they won’t, because being “open” is their first priority.

This demonstrates a colossal disconnect between the people who work at Google and the average American consumer.

Google is somewhat infamous in the industry for having some really bizarre hiring practices. They once advertised a recruiting site with a billboard that urged Bay Area motorists to go to a Web site with the address “the first ten-digit prime number found in consecutive digits of e … dot com.” Going to that Web site — some mathematical savant figured out the ten-digit numerical address — takes you to another numerical puzzle, this one even more mind-boggling. Only by solving this second riddle could you get to Google’s actual human-resources site.

The company earned a lot of respect in the industry for being so balls-out elitist in its recruiting practices. It’s their prerogative, after all. They’re looking for people with multiple doctorates in abstract numerical theory; why not weed out the obviously underqualified candidates before they ever get your phone number? It makes sense, if a sort of twisted, malicious sense.

But if you’re trying to design a product, something for everyday people to use, you do not want a multiple-PhD in finite analysis calling the shots. You want somebody who goes home at the end of the day and talks with his kids and plays with his dog and has sex with his wife. You want somebody whose personal priorities are, if not precisely typical, at least similar enough to the priorities of your average potential customer than he can understand what they’ll like and dislike.

Google’s apparently put a lot of smarts into this whole Android thing. It’s a pity they’ve put all the wrong kinds of smarts into it.

Aaaaand now my friend who said I’ve been writing too much about technology is royally pissed at me. Well, shit. It’s been that kind of day.

• Stuff happens when you’re not looking
• There aren’t any
• I guess there really aren’t any new ideas