Computers are cars

How often do you buy a car? Some drivers turn in their cars every couple of years and get a new model. Others drive them until the wheels come off. Others split the difference, trading in every five or six years.

Generally speaking, people change cars because something wrong with the car that they have, or because their requirements have changed, or because they just want a different one.

Something’s wrong with your car: You have an accident, so you’re in the market. Or your car needs major repairs that seem too expensive, given the value of the car. Perhaps you have reasons to stop trusting its reliability, and don’t want to be stranded.

Your requirements have changed: You have a new baby, and so a two-seater won’t work. A new hobby means getting a sports-utility vehicle. A new job’s commute calls out for something with better fuel efficiency. A teen driver means adding a vehicle.

You just want one: A bonus lets you buy the sports car you’ve always wanted. You want to make a statement by driving a hybrid, or by getting a Lexus or BMW. Your brother-in-law got a new truck, and so you want one too.

What about computers? People used to replace their computers because the new computer had essential features that the older computer’s didn’t have. It was enough faster to let you do more tasks. It had better connectivity. It supported more storage options.

The problem is that today’s computer hardware and software is so darned good that the regular upgrade cycle has collapsed.

Look at Microsoft’s terrible financial results, reported on July 30. The company announced that its revenue declined 17% from the same quarter of the previous year. You might say that’s because Microsoft is having its butt kicked by companies like Google and Microsoft. That’s part of the situations, sure. But that’s not why Microsoft, and other computer companies, are in such a funk.

Consumers and businesses used to buy lots of new “stuff” every year or two or three. New computers, going from a 286 to a 386 to a Pentium. New modems, from 1200 bps to 9600 bps to 19,200 bps. New printers. New flat-screen monitors. New scanners. New operating systems. New productivity suites. The changes in hardware, in operating systems and in major applications was so significant that the upgrade was perceived to be a good value.

Today, consumers buy computers because something wrong with the computer that they have, or because their requirements have changed, or because they just want a different one. Businesses only buy new desktops and notebooks if they absolutely must do so.

Apple and Hewlett-Packard have done a great job of tapping into the “just want one” segment of the computer market. Apple has positioned its iMacs and MacBooks as cool and sexy. HP has brought out compelling innovation with its TouchSmart desktop and cutting-edge design. That’s why they’re succeeding.

Other companies, from Microsoft to Intel, from Dell to Lenovo, aren’t succeeded in convincing consumers that they need new hardware or software. Got a 2.1GHz Core 2 Duo processor in your computer? That’s fast enough. Got Windows XP? That’s good enough. Got Office 2003? It has every feature you need. A quad-core chip, Windows Vista, Office 2007, aren’t going to move you… or move your credit card.

We’ll have to see if Windows 7, Office 2010 and eight-core chips can jolt consumers into wanting to upgrade. If not, be prepared for more sad earnings from the PC companies.

Google’s Chrome OS is less than it looks

My email inbox was filled this morning with questions and opinions after news came out about Google’s new Chrome OS.

Was this another frontal assault on Microsoft by the Googleplex? Is Google becoming too big and powerful? What’s going on?

My feeling is that there’s less to this announcement than breathless pundits are panting. Nobody should be surprised that Google wants to enable the development of standalone Internet devices. Chrome OS is an obvious extension of what the company is already doing with the Chrome browser and Android mobile platform.

Think about the classic functions of an operating system. It’s a program loader. It provides an abstraction layer for hardware. It allocates shared hardware resources, like processor time, memory and storage. It manages IO devices, including the focus of input devices. It provides a consistent user interface. It provides security between different processes. Stuff like that.

If your focus is to provide Internet access for the masses, using devices like netbooks, nettops and handhelds, you don’t need the heft of full-featured, industrial-strength operating systems like Windows XP, Windows 7 or Mac OS X. You want something thinner.

The best thin operating system for smaller Internet-centric devices is Moblin. However, Moblin has one serious flaw: the project is driven by Intel, and the company’s primary motivation is to get the market to adopt its Atom processors. Thus Moblin is designed to freeze out competing chips, such as anything based on ARM. Until Moblin (recently put under the auspices of the Linux Foundation) becomes truly hardware-agnostic, its potential is limited.

That brings us to Chrome OS. Is it a competitor to Windows or Mac OS X? I don’t see it. What I do see is that Chrome, when it appears, will be essentially a delivery platform for browser-based computing using the Chrome browser.

Currently, Chrome is in the second tier of browsers, along with Safari and Opera. Chrome OS will move it into the first tier. Software developers must begin making sure that their applications run cleanly in the Chrome browser – just like they must do today with Internet Explorer.

And that, my friends, is all that this announcement means.