Prototypes are hastily assembled contraptions for trying new ideas on, and nobody cares how stable or reliable they are as long as they run long enough to see the idea in motion. The prototype's complete lack of stability stays with it like a teenager's clumsiness until, after a maturation process, the maker ships what it calls a finished product. You'll see that teenager in every desktop computer you can buy today, because even after two decades the PC is still just a prototype, and nobody cares if they go up in flames every now and then.

It shouldn't come as a surprise that most PC owners don't heed the warning not to run pre-release software. The fact is that pre-releases are popular and used as much as the so-called Final versions of the same products. Many PC users think nothing of using a "Beta" instead of an earlier formal version. The association of the word Beta with new ideas is so strong that some products never emerge from this arbitrary label, so good is it for marketing. (Reader responses)
Reader responses to this paragraph

1. Articles By James Pinto (referals: 9)
2. Bachus (referals: 7)

Want this paragraph to link to your response?

Close this

It's important to understand that this smearing of the line between informal prototypes and formal releases isn't to do with a dissatisfaction with the overall quality of software -- as if people are saying they can't tell the difference between an untested program and one that's advertised as finished. Indeed, what one company calls finished software is often as bad as another company's proof-of-concept. But as long as it is stable enough then that's all the user really wants. If you have any doubts just look at people who use Palm organizers - the reset pin is so critical it's now built into the stylus.

The purpose of a PC gets completely re-invented every couple of years. Lets say you start in the late 70s with the Apple II and Visicalc, which re-invented computers as a desktop accounting accessory. In the mid 80s it was re-invented as a desktop publishing tool. In the late 80s it was used for groupware such as Lotus Notes. In the 90s it was re-invented as a device to send and recieve e-mail, then a device for browsing the web, then a device for trading music. In 2000 it was re-invented as a major component of a home entertainment system. And now, in 2001, it became a tool for maintaining a public diary.

With re-invention coming so fast, how could there be enough time for the PC to mature? Once you've debugged everything for one application, you find that everybody has moved on to something else.

Unfinished software was shipped regularly all through the PC era, but to nobody's detriment. They were installed by users when they worked well enough, not when they were perfect. And ever since the Internet made distribution costs so negligable, everybody has subscribed to the Bazaar-model's mantra of "Release early, release often."

Nobody ran the "Final" versions of Netscape or Napster because it was more important to have them now than wait until the nice programmers had eliminated all those nasty bugs. You can live with a few crashes. Nobody dies when a PC goes down. You just reboot it or re-install the software. That's all. And despite the obvious stress, hardly anyone is changing their ways.

PC owners aren't learning to accept unreliability as the norm, they aren't "being conditioned by Microsoft to enjoy crap", they're just reluctant to wait for the next better mouse-trap. And if you're a stickler for reliability, being dragged kicking and screaming into another crazy fad, then gee, we guess that's real tough, ain't it?

This could understandably piss off those who reckoned their ultra-stable alternative operating systems and platforms would crush Windows and Intel. But is WinTel popular because of dirty licensing arrangements, or because everything else was always two steps behind? Windows was happy to cater to impatient gadget lovers when other platforms didn't, even if it meant shipping unfinished software and dodgy-but-prototype-friendly development tools such as Visual Basic. Yesterday the punters were buying computers to write e-mail to Grandma, today they're buying them to show digital photographs, tomorrow they will be adding a CD burner. Hells bells, what a pace.

That means introducing new hardware and software that's untested in the real world, which also means the inevitable re-allocation of the developer's resources away from bringing the old technology to maturity. The teenager picks up the skateboard before he finishes learning the stunt bike.

The non-fatality of a computer crash has turned us into desktop adventurers. Since everything a computer knows can be copied to somewhere safe, and every component can be replaced on the cheap, where's the risk? Think about how your game playing strategy changes when you discover the cheat for infinate lives. Instead of playing cautiously, you go off to explore. It doesn't matter if the avatar 'dies'.

So if technology needs decades to mature, what would it mean if every PC really was as rock-solid and stable as so many vendors love to claim?

It'd be a great opportunity for someone to mess it all up again.