gavinshearer.com

« Mac Mini, Unwrapped | Main | Mac Mini: Hands On »

January 20, 2005

Moving The Industry Forward

I wanted to expand a little bit on last night's post about the Mac Mini. In it, I complained that "...some modern PC laptops don't come with WiFi, Bluetooth, FireWire, DVD drives, or even modern (2.0-era) USB" which is "...bad for the industry, because it retards technological progress in general."

A few of my friends have asked: what the heck am I talking about?

Personal computers aren't "for" anything in the same way that other, physical objects are. The last 30 years of personal computing have been, quite literally, a series of people figuring out neat things they can do with their machines and then getting their friends to do it, too. BOOM! Overnight industry.

Spreadsheets? Dan Bricklin was a Harvard MBA who got sick of "running the numbers" by hand. Desktop publishing? Paul Brainerd thought that computer documents ought to work more like a sheet of paper. Marc Andressen had a jones to look at images and text, together, on this geeky thing he and his friends were screwing around with at college. Shawn Fanning liked the idea of sharing his music with his friends.

In each case, these innovations depended on that which had come before. The spreadsheet needed a non-mainframe computer. PageMaker needed that, plus a GUI. Mosaic both of those, plus the Internet, HTTP and HTML. And Napster required everything of the other three, plus digital music (which itself required CD-ROM drives and sound cards). Good, cheap, widely-available modern hardware is the precursor to software innovation.

Now consider the modern PC. Many Pentium 4 machines out there still ship with freakin' parallel ports. In human terms, this is like having a prehensile tail as well as an opposable thumb. It's just embarrassing.

Why the legacy equipment? Because the PC hardware business is a low-margin, high-volume business that's based on "shipping boxes" with little regard for the future impact of that box. PCs are usually bought based on comparative specifications (e.g., Box A has 600mhz processor and 40 GB hard drive, Box B has a 500 mhz processor and 30 GB hard drive. Hence, Box A is better), so manufacturers tailor their specs to what their competitors are offering. Innovation proceeds along grooved tracks - processors get faster, drives get bigger, CD-ROM drives get faster, etc. - but no real time or money is proactively spent by the PC makers to introduce new kinds of hardware or features that will obviate the existing stuff. Status quo = profit.

Take USB. USB is, without question, a terrific peripheral technology. In the old days, your computer would have a variety of ports that would allow it to talk to external devices. These ports were "dumb" (no ability to auto-detect whether something was plugged in to them), and slow (say, 57,600 bits per second or so), but they were compatible with the growing body of devices (printers, scanners, mice, keyboards, etc.) that were being introduced by third-party manufacturers. Consumers hated having all these little ports on their machines because they were complex and prone to trouble (if anyone remembers DOS or Windows 3.1, you also remember the fun of IRQ conflicts). These ports were also one-at-a-time; if you had a printer plugged in and wanted to use a modem that required the same port, you'd have to unplug the printer to use the modem. And vice versa. This "pain in the ass" factor tended to impede people's adoption of both printers and modems (or more than one peripheral in general), because of the issues changing out the devices.

So USB promised to change all this. It's fast (12,000,000 bits per second), smart (it detects whether something is plugged in or not), daisy-chainable (you can support up to 127 devices off one USB bus - just add a hub if you need more ports), didn't experience IRQ conflicts, and even had power - devices could get their electricity from the USB bus itself, and no longer needed to be plugged in to the wall. USB was even invented by Intel, so it was backed by one of the most powerful and respected companies in the industry.

USB, on introduction, flopped.

No, really. PC manufacturers at first didn't include it on their designs. This was a combination of the cost and education issues. Consumers "weren't asking for it", so makers didn't include it. Of course, consumers "weren't asking for it" because they didn't know about it ... because their PC didn't come with it. Chicken and egg.

Then Intel twisted some arms, and a few manufacturers (Compaq, for instance) started including it. Still, nobody used it because all the keyboard, scanner, mouse, and printer vendors were making devices compatible with the "larger market" - the old, serial- and parallel-port using public. Again, chicken and egg.

What USB needed - and ultimately got - was a shot in the arm. And the shot in the arm came when Apple introduced the iMac in 1998.

The iMac was watershed for a lot of reasons, but the most-overlooked, IMHO, was that it jumpstarted the USB market. See, the iMac didn't allow any kind of peripheral to connect except for USB. The industry snickered - didn't Apple know that there weren't any USB peripheral vendors out there? - but, overnight these USB peripheral makers had a captive market for their stuff. And so they started making it, and shipping it, and the iMac's success meant that millions of people were buying USB stuff and loving it.

Time passes. A few years later, the penetration of USB peripherals was sufficient to make its other advantages obvious to the rest of the world, and the PC industry started pushing USB. And today, the "USB economy" - keyboards and mice, of course, but also USB flash drives and iPod Shuffles - is huge.

The exact same thing happened with WiFi technology - it'd been around for a long time, but it was expensive and exotic. Apple shipped the iBook with WiFi, shipped a base station to go with it, and now WiFi is everywhere. In fact, the entire Internet and telecom industries are undergoing some major change, and a lot of it is being driven by cheap, reliable, wireless technology at the PC level.

My point in all this is not to praise Apple, per se (they did a lot, but they did some of it from necessity), but to instead damn the rest of the PC box makers for lacking anything that even remotely approximates vision. (Dell is one of the worst, here.) By "giving the people what they want" (which, mysteriously, is almost exactly what they've come to know), these companies are effectively freeze-drying our technical infrastructure at the 1981 IBM-PC era. They're killing tomorrow to eke out a slightly-more-marginal profit today. Golden goose, anyone? Hello?

Microsoft is spending a lot of time and money on designing the "PC of the future", and I commend that. My suspicion, however, is that Microsoft is still stuck lobbying the Toshibas of the world to produce radical new hardware designs (e.g., Tablet PC) for customers. Some of this lobbying is successful, some not. Some markets will take longer to mature than others. Hardware makers, however, generally don't care about long-term industry growth. They're too focused on short-term earnings.

So maybe Microsoft should get in to the PC business. It could produce the high-end "reference" machine, and let the other manufacturers follow suit and fill the needs of their respective niches. Hm.

In the meantime, Dell is happy to sell you a $400 PC that's the best 1981 had to offer. But don't let anyone tell you it's the equivalent of the Mac mini.

Posted by Gavin Shearer at January 20, 2005 4:13 PM. Posted to Geek.

Comments

Post a comment

« Mac Mini, Unwrapped | Main | Mac Mini: Hands On »