February 2002

One night at the beginning of the O'Reilly Open Source Convention (OSCON) last July I had dinner with three of my best friends: Craig Burton, Eric Raymond and Dave Winer. Prior to that day, none of the them had met in physical space, and I was a little concerned, since all three were both powerful personalities and very frank people who had disagreed publicly on various matters -- Dave and Eric especially. But I was also excited. Common interests had begun to outweigh differences. Eric, for example, had taken the lead in getting XML-RPC (an open source effort led by Dave and his company, Userland) support baked into Python. Meanwhile, Dave and Craig had both become increasingly vocal advocates of Eric's cause -- Open Source -- to their native professional worlds: commercial software development and enterprise computing.

The dinner went well. We talked about "inclusion"and how it applied to both developers and customers, which were the separate constituencies of the week's two conferences: Open source developers were in town for the O'Reilly Open Source Convention, while commercial software customers -- mostly large enterprise IT managers -- were in town for the Burton Group Catalyst Conference. (Background: The Burton Group was the eponymous creation of Craig Burton in 1989. He left a few years later.)

Would this be an "inclusive" week? I wondered.

The cultural gulf between these two conference's attendees was as wide as the one between golfers and gamers. But I could see the distance shrink at the table. Dave had been chosen (at Craig's recommendation) as the closing speaker at Catalyst (following Dave Thompson of Microsoft) precisely because he could speak as an independent developer with code-level connections to both the open source community and Microsoft (with which Dave and Userland co-developed SOAP and co-initiated XML-RPC). Craig also knew Dave wouldn't say the diplomatic thing. Sparks were likely to fly.

But where would they land? Would anything catch fire? I hoped so.

That dinner came at the close of Day One at Catalyst. That morning I arrived early for Burton Group president Jamie Lewis' agenda-setting keynote, and caught the end of the preceding opener, "Managing Information Technology for Competitive Impact" -- by Craig Watson, the CIO of FMC -- guys known for making tanks for the military and other very large professional toys. It was anesthesia. The guy droned about "penetrating customer value chains," and other stuff, all in TechnoLatin -- a language to which I'm so allergic that I didn't even want to understand it.

Jamie's talk was about infrastructure: "I before E" was the title. One item stands out in my notes, though it may not be a verbatim quote: Customers, big ones, like big vendors because they are stable. Microsoft is a big stable company that other big stable companies expect to make a good partner for building solutions and pushing them out into the marketplace.

Didn't we used to say that about IBM? And isn't IBM still better positioned to do exactly that? Well, not if Microsoft is taking the obvious lead in building business infrastructure on the Net -- or, more precisely the .Net.

Clearly Microsoft has been doing a good job of making itself central to the Catalyst audience's stated concern: building commercial enterprises on network infrastructure. I just wish TBG's language weren't so generic and impenetrable.

The next speaker was another Burton Group guy who had this to say about something that (trust me) matters:

Directory serivices are information designed to mitigate the congintive costs associated with community membership and to help coordinate behavior among community members.

Hah?

Later the same guy said P2P was "an admission that we are not alone in the world." (Yeah, and so is watching TV in your hotel.) Then he said this had something to do with "transactions." If I understood him right, he believed everything had something to do with transactions. But if that were so, our most meaningful relationships would be with vending machines.

Bazaar Behavior

Events like Catalyst and OSCON are industrial bazaars: markets that materialize for a few days in physical space, with booths not too different in purpose from those operated by merchants at trade route crossroads for the last several millenia of human history. With this observation in mind, I began to think about the featured conflicts between open source and Microsoft folks, especially as they relate to customers in the bazaar where vendors from both sides sell their wares.

While other folks took the stage at Catalyst, I sat in the audience and began to wonder about how well -- and how poorly -- both sides understood the bazaars they shared. After awhile I began to wonder: might it be possible that the open source folks know more about The Bazaar — the most original and fitting synonym for The Marketplace — than the greatest marketing organization computing has ever known?

I decided the answer is a qualified yes -- and that the open source community needs to embrace and extend the most precious thing that Microsoft has and Open Source doesn't (at least not in huge numbers), which is paying customers. Why? Because it's their hearts and minds we need to win, and Microsoft hardly ignores them, it is busy mischaracterizing them, and that may open a window of opportunity.

Craig Mundie (quoted at length in my last report) speaks from a point of view is not anchored with customers, but with the commercial software industry. Big difference. Yes, Mundie is right that commercial software companies face a choice among many different business models and licensing schemes, including the GPL. But the "debate"Mundie wants is not just about business models. Nor is it the exclusive concern of Supply. It's also about Demand. But Microsoft doesn't know that -- not yet, anyway.

To Microsoft, this "debate" is between two kinds of Supply: one "shared" and one "open". But there is a much bigger context here, and that's the relationship between Supply and Demand. Microsoft — like all industrial giants — naturally wants to "own" demand. To dominate it. To control it. We see this not only in the company's aggressive business practices, but in documents like the Passport home page and the Hailstorm White Paper. The Passport page is a document called "Consumer.asp". The Hailstorm paper not only divides the world into "users" and "businesses", but positions users as helpless victims of technology.

This is in character with Microsoft's nature as an industrial producer serving billions of consumers. It is also, in respect to actual nature of its customers and users -- from Fortune 500 companies to individuals using home computers -- a weakness.

Some context.

Ever since Industry won the Industrial Revolution, Supply has been in command of its relationship with Demand — or at least the ammassed form of Demand we call "consumers". To measure the degree of tilt in the power balance between large producers and "mass markets" of consumeers, consider this one fact: Before John Wanamaker invented it in the late 1800s, there was no price tag. Why? Because price tags were anathema to the nature of relationships in real markets.

Before the Industrial Age, markets really were markets, which we also called bazaars.

The Industrial Age was not about bazaars. It was about production and distribution. What industry called "markets" were just abstractions: synonyms for populations, forces, desires and behaviors at the far ends of vast distribution deltas that fanned out from a relatively small number of large producers to millions or billions of consumers. What we call "consumerism" today is actually producerism: a relationship between the largest producers and billions of consumers that is roughly the same as that between ants and aphids. The name for that relationship is "marketing."

Microsoft is exceptionally good at marketing. In spite of its constant talk about "competitiveness" and "innovation," what Microsoft does best is pay attention to users' wants, behaviors and appetites: Marketing 101. Even if Microsoft had not been blessed with competitors who flat-out sucked at marketing, the company would be an exemplar of the practice.

The problem with marketing is that it has always operated in a world where producers were in charge. Supply ruled Demand.

Now, as markets get more networked and relationships get more personal, the balance of power is shifting back to Demand. Consumers are turning into customers again. They are no longer what Jerry Michalski calls "gullets who live only to gulp products and crap cash." They have more than money and choice. They have experience, good ideas, and the ability to talk about both with whomever they wish. This makes them a lot harder to please, but also a lot more valuable. Same goes for everybody else contributing to a market's conversation. If producerist industrial creatures don't adapt, they'll fail.

But the producerist mentality dies hard. Again, look at how "user" and "consumer" are employed in the Hailstorm White Paper. Look at what is implied by the term "experience". The language may be concerned and respectful; but it's still clear Microsoft is farming aphids here:

Over the past 25 years, information technology has provided incredible value for users and businesses. Even so, there is room for significant improvement. Individual applications and devices are narrowly focused and ignore the rest of our world. People today are forced to adapt to their technology, when the technology should be adapting to them.

People are frustrated and confused. Sometimes it seems that every program, every Web site, every device has its own set of rules. For instance, entering a friend’s new phone number into your PC requires a particular sequence of keystrokes and mouse clicks.  But to enter that same information into your Palm Pilot, Pocket PC, or cell phone, you have to learn a completely different interface – right down to relearning how to draw the letters of the alphabet!

People are not in control of the technology that surrounds them. We have important data and personal information scattered in hundreds of places across the technology landscape, locked away in applications, product registration databases, cookies, and Web site user tracking databases.  Having a friend’s phone number programmed into your cell phone doesn’t make the number available to your PC-based email application: those two technologies are unable talk to each other.

This is extremely condescending stuff that's easy to mock and dismiss. But the better response is a strategic one. Because what we have here is an opportunity to side with the bazaar — to treat "consumers" as what they really are, which is customers with a lot more intelligence than Microsoft gives them credit for in this paper.

It won't be easy, because IT customers understand software in largely commercial terms. In other words, their conversation is in much closer alignment with what Craig Mundie says about, say, "intellectual property" than with what Richard Stallman says about the same subject. IT customers are also -- by demographics if not by definition -- Microsoft customers. They're almost pickled in the stuff.

While Eric Raymond might be right when he says "software is largely a service industry operating under the persistent but unfounded delusion that it is a manufacturing industry," that delusion is still common wisdom in IT business -- on both the Supply and Demand sides of countless relationships. Manufactured or not, software is a massive business category.

What we have to do is change that category by plugging customer needs directly into open source development. Will that corrupt open source development? Not if the purpose isn't to "own" the customer or otherwise restrict the customer's choices.

The trick is to replace controlling relationships with cooperative ones, based on trust, openness, identification with shared interests and constant conversation. Both the free software and open source communities are good at that. Microsoft isnt -- at least not yet.

Code

Indeed, the "licensing structures" Mundie talks about conceive code as capital: software products are manufactured goods. Concepts like "intellectual property" are easy to understand and argue about as long as one continues to conceive of that property in material terms.

Yet code is not material, and no amount of lawmaking or marketing can make it material. But we've been wrestling with this fact for decades now. Are we making any progress?

I'm beginning to think that the deepest fact in this matter is not that software wants to be free, but that code wants to be public. Meanwhile too many of our laws and business practices cannot comprehend this simple idea. It's too far below their immediate concerns. Like the core of the Earth, it's nice to have but too deep to appreciate.

One big problem for Microsoft is that it lives in a world increasingly built with public code that oozes like lava out of the relatively free and open ground below everybody. We're making a whole new world here, but we're doing it together. And that includes Microsoft, which in fact does contribute to common infrastructure. SOAP is a good example.

It's plain that much good has been produced by what Mundie calls the "commercial software model" or it wouldn't have customers. But the Net on which all of business increasingly relies is at most only a partial product of that model. It wouldn't be here hackers were not committed to building infrastructure with public code.

What hackers know and Craig Mundie doesn't (yet), is that there is much to the nature of code that can neither be comprehended nor represented by the conceptual system Mundie employs. At a deeper level -- the nature of software itself -- the principles of business don't apply for the same reason that the principles of mechanics don't apply to chemistry, even while mechanics depends on chemistry as a deeper principle. (You'd rather make a clock out of iron than sodium, but neither element has a thing to say about telling time.)

Business needs more of that public infrastructure Jamie Lewis has been talking about. It needs programs, operating systems, device drivers, file formats, directories, security models and protocols that everybody can use because they are public, which is to say both accessible and uncomplicated by an owner's hegemonizing agenda.

The difficult challenge for the free software and open source folks will be figuring out how to embrace customers without constantly dissing Microsoft. The temptation to oppose Microsoft constantly is more than tempting. It's standard. "It takes two to tango," one commercial developer wrote to me a few weeks ago. "If Microsoft ever chose an enemy who was willing to share the cursor, they would finally have met their match. It hasn't happened yet. Your friends are still saying 'It's All About Us' which is complete bullshit. They never talk about anyone but Microsoft and themselves."

One thing is clear to me while I sit here at Catalyst: it needs to happen. But can the open source folks make it happen? That's what I'm hoping to have answered across town at OSCON.

Doc Searls is Senior Editor of Linux Journal. His monthly column is Linux For Suits. He is also a co-author of The Cluetrain Manifesto.