About this site
















The Micro Channel Architecture

Friday, July 6, 2001 by Dave Winer.

The Micro Channel Architecture Permalink to The Micro Channel Architecture

Let's go back in time to the PC industry of 1987.

Things were moving really fast. In the late 70s and early 80s Apple had risen to the top of the early personal computer vendors. Steve Jobs was a cultural icon, deservedly, on covers of so many magazines. He told a revolutionary story -- computers could be human-scale tools, not just data processing systems for large corporations. He described the Apple II as a fractional-horsepower computer, and hit the nail on the head.

In 1981 IBM introduced the PC and took the leadership from Apple. Looking back I think the reason it worked so well is simply that the PC had much more memory than Apple's computer.

Then came Compaq and then other cloners. Where IBM was slow and lumbering and arrogant, Compaq sold us the products we wanted, and forced IBM to climb a very uncomfortable curve. IBM could see the future clearly, and didn't like it. The inexpensive fractional-horsepower desktop computers were quickly becoming as powerful as the multi-million dollar mainframes that were IBM's mainstay.

In 1987 IBM made their move, releasing a three-way noose for the nascent independent computer industry lead by Compaq. The three components were the PS/2, OS/2 and the secret sauce -- the Micro Channel Architecture.

What a beautiful crock it was. I'm no expert on bus architectures. Probably one person in a million understood what it was, and no one had a clue why it was so killer (it wasn't) and that's why it was so effective at clouding the judgment of so many. We simply didn't understand what they were talking about.

Compaq, with its death designed, responded with EISA, an open equivalent of the MCA. We saw their moves as pathetic death throes, but it worked! Today the Micro Channel Architecture is a historic curiosity. IBM guessed wrong, we loved their PC, but when IBM wanted to kill it, we said no.

Zoom forward in time Permalink to Zoom forward in time

In 1993 everyone thought Bill Gates knew where the software industry was going. If you stood up and said "Some random person you've never heard of is going to define where this industry is going," everyone at Stewart's or Esther's would have laughed, what a silly idea, that's not how the world really works.

If you're shaking your head in agreement, think again. The "random person" was an amalgam of Tim Berners-Lee, Jim Clarke and Marc Andreessen. I wonder if any of them dared to believe, as Steve Jobs clearly did, that they were playing such a pivotal role in technology. As the Web grew it killed every advanced project at Microsoft. None of the people who anticipated the Web worked at Microsoft, if they had, there's no doubt Microsoft would have killed the idea -- it was too open, too fair, and it undermined all their plans.

You could visibly see them figure this out at Microsoft in 1994 and 1995, and now in 2001 the Web is as much Microsoft's as the PC was IBM's in 1987. And now they're making the same fundamental mistake that IBM made, they're going to try to foreclose on the Web, all the annoyance and disconcert will be organized to feed their machine, according to the plan.

The killer triumvirate Permalink to The killer triumvirate

Microsoft's killer triumvirate is Passport, .NET and the secret sauce -- HailStorm. It's a total lockdown, just what IBM proposed after quelling the Apple-lead insurrection as they tried to dispose of Compaq and the cloners. This time what's at stake is the locked-in-ness of Windows and Office. I think Microsoft guessed wrong. We love the Web, now that Microsoft wants to kill it, we will say no.

It's so clear, what comes next is the open equivalent of what Microsoft is proposing, as EISA was the open version of the MCA. The art of membership is clear, it's not rocket science, and it's not worth giving Microsoft a lock-in. .NET is merely a runtime environment and a set of development tools, not unlike Zope, Java or my own Frontier. Since .NET supports SOAP and XML-RPC, we can implement interfaces to the EISA-of-the-Web in Microsoft's environment.

There appears to be nothing controversial about .NET. But HailStorm -- that's the mystery. Where I didn't know much about bus archtitectures in 1987, I know lots about schema for everyday data, calendars, business cards, mail messages, weblogs, spreadsheets, outlines, etc. I bet that the schema of HailStorm are going have poison pills, probably patents, all kinds of non-disclosures, but the technology is thin. We'll have to fight them the way Compaq fought IBM. And unless there's something I'm missing, Microsoft will hit the same wall that IBM did. Their motto of 2001 might well be Who wants to be locked in the trunk today?

The 1990s were a great period of growth for Microsoft, but they may have missed the message. We want our independence. The users will want that too. As the tie-ins in Windows XP become evident, there will be more disaster stories as the full effect of Microsoft's lockdown is felt by users. They'll look for other ways to get what they want. Let's make sure they have choice.

Dave Winer

© Copyright 1994-2004 Dave Winer. Last update: 2/5/07; 10:50:05 AM Pacific. "There's no time like now."