Open vs. closed networks

As part of my [thesis][] research, I have been reading a stack of books, including [Linked][linked], [Six Degrees][sixdeg], [Design For Community][dfc] and [The Future of Ideas][tfoi]. At the same time I’ve been exploring and studying various online communities, trying to determine what makes them tick.

Is there a common characteristic of successful communities? So far, I’ve been nursing along the notion that “openness” is a common feature (among others) of the groups and communities that I’ve been studying.

[thesis]: /archives/2004/05/my_thesis_topic.html

By “open”, in this context, I don’t mean source code or software licensing, but participation. An open community is one where anyone is invited to join, provided one supports the goals of the group and adheres to the established rules of conduct. To put it another way, the barrier of entry to join an open community is very, very low.

So, last night I’m reading chapter eight of Linked (a terrific book by the way), wherein a number of connections are made between the mathematical properties of man-made networks and naturally occurring phenomenon.

Up to this point, the network models presented overwhelming favor the earliest entrants into a new industry (“the oldest nodes win” in network-speak). By way of explaining the success of Google, a late-entrant into the search-engine market, [Barabási][alb] injects the notion of “fitness” into his model, to measure the relative competitive quality of each node. Simple put, the superiority of Google’s technology, its fitness, gives it a competitive edge that outweighs the first-mover advantage then enjoyed by the previously-established players.

All of this takes a decidely unexpected turn, however, when Barabási draws a parallel between the quantum theory that predicts [Bose-Einstein condensation][bec] — and the market dominance of Microsoft. In essence, just as a certain rare set of conditions in a network of sub-atomic particles can create a highly-condensed new form of matter, a similar set of circumstances can give rise to a monoply in an economic network. Barabási describes it as the “winner takes all” scenario.

But what factors led to the dominance of Microsoft, and more specifically Windows? When it debuted, Windows was neither the first-mover nor the fittest competitor in the graphical operating system market. That spot was held by Apple and the Macintosh. So what was it about Windows that enabled Microsoft to replace Apple as the market leader?

I propose that “openness” had something to do with it. Specifically openness with regard to hardware manufacturers. While no one but Apple was allowed to make hardware that ran the Macintosh operating system, any [kid in a dorm room][dell] was able to make computers to run Microsoft’s OS.

I know that the idea that Microsoft benefitted (greatly) by becoming the de-facto standard operating system during the PC boom in the early 80’s is not a revelation. What I have not seen, yet, though is this idea captured in the context of network theory. I am neither a physicist nor a mathematician but it occurs to me that in the same way “fitness” was added to the model to explain Google’s success, a measure of “openness” could be injected to describe Microsoft’s success.

In the language of network science, openness would be a measure of the “cost” required to link to a node. In the case of hardware manufacturers, “linking” to either Apple or Microsoft is equivalent to creating products that are compatible with the Macintosh or Windows operating system. Viewed from this perspective, the cost to “link” to Apple was very, very high (approaching infinity), while the cost of “linking” to Microsoft was very, very low. As a result, despite being a late entrant to the market, and being an inferior product (at least at that time), Windows nevertheless quickly became the dominating force in the industry.

Looking beyond the operating system market, what other examples exist where the “openness” of a network gives it a distinct competitive advantage? A few examples jump to mind. Captialism vs. Communism. Democracy vs. Monarchy. The Web vs. AOL/CompuServe.

From a software developer standpoint, it also explains the advantage “simple” standards have over more complex, yet full-featured standards. For example XML-RPC vs. SOAP. The complexity increases the “cost” associated with supporting the standard, and hence the barrier to entry is higher for one rather than the other. It also shows how that barrier can be lowered via better development tools and supporting standards like WSDL and UDDI.

It is also interesting to consider that the biggest threat to Microsoft’s dominance is Linux, an operating system that is arguably infrerior (from a average user’s perspective) yet is much more open.

This model seems so simple and perhaps even obvious, that I expect I may not be the first person to happen upon it. But so far I have not run across this idea in my research.

So what do you think? Am I barking up the wrong tree? Is this an already well-worn path in the study of network systems? Or is this, in fact, a novel observation?



One thought on “Open vs. closed networks

  1. Hi Mason! long time no see. howdy! 🙂

    A couple of feedback on this:
    1. MS versus Apple. It is a matter of marketing. Microsoft, most likely intended, otherwise evolved into, targeting at the majority population. Apple, on the other hand, targeting at only a certain segment. Like Harley Davidson, there is a pride in both the maker and the user of it. It really has nothing to do with the openness of the product. A product is made by people, and used by people. Apple is sold only to users who have natural connection with the maker. Microsoft, on the other hand, is willing to and has been successfully building a bridge: a bridge between the maker and the user. Openness is a component of this bridge.
    So goes with Linux. It is completely “open”. Yet it is “closed” to the majority. I smell rebellion and extremist from it. True linux passionates don’t like to install with package, what they really enjoy is, compiling everything from source. For this reason, I don’t think Linux is ever going to be successful in the consumer market. Unless, someone, builds a bridge.

    2. Online open network and close network. I have just recently got introduced to, a close-loop network. Comparing to an “open networking” where anyone can just sign up a free account; a close-loop network requires a newcomer to be introduced by someone who is already in the network. Trust is an interesting topic here. Any virtual community, any online network, will eventually carry the same characteristic as the people community and people relationship. Trust is very important to the virtual world. Close network is an attempt to address this problem. The technology is not mature yet, but it is promising.

    It’s late, got to go. this is a topic I have been reading. Please keep me posted and I would love to share more later!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s