This article explores such questions as what happens as new
actors come on board with respect to the
Six
Degrees of Kevin Bacon game? Naïvely, one would
assume that eventually a seventh degree would be required.
What happens in practice is that
power
law effects kick in. Most new actors end up being
supporting actors in a movie with a popular actor, and in the
process the popular actor gets more popular.
What are the implications? Pretty profound.
For decades the presumtion has been that radio spectrum is
scarse and that organizations like the FCC are required to regulate
them. Recently this
has
been questioned. Perhaps adding stations increases
capacity. Counter-intuitive, perhaps, but not impossible.
Now lets turn to the realm of people. Clay Shirkey has
ruminated on the relationship between
Communities,
Audiences, and Scale. What he says seems reasonable, but
the question remains:
What if Clay Shirkey is wrong?
David Reed is the main proponent of this concept -an interesting guy.
Bandwidth goes up for the same reason that bandwidth in a room increases if everyone talks to their near neighbours rather than take turns shouting across the room.
Unfortunately, latency, routing and probability of packets getting through take a pounding, and you are also putting your own battery to work on behalf of others, routing their packets. You gain by being unco-operative, though the MANET community gains though cooperation.
Some of the more modern protocols do take power into account, along with things like node mobility, past history of connection quality, and route better, but it is still research. Interesting research though -something you can do at home with others in your metropolitan scale ad-hoc network.
interesting projects in this area whose URLs escape me: MIT GRID project, something in dublin, INRIA.
Posted by Steve Loughran at
Steve, it is the concept of 'neighbors' itself that may need to be rethought. You and I met at a Chris Sell's devcon. We work for different companies and are physically separated by a few time zones and a few thousand miles. We have worked on some of the same projects, but rarely at the same time.
Yet there is some very real sense that we are 'neighbors'. And we don't even need to shout across the room to talk.
I'm not exactly sure why this is, but in my case once I started working in open source, the number of neighbors I enjoy went up immensely. And since I started blogging, it has literally exploded.
It's because the way people connect and interact in net communities (such as open-source work and the blogosphere). In the real world, competitive differences between companies stops most attempts at interoperability. Online though, most useful protocols and applications (the things we need to get anything done) are done by large groups of volunteers, who tend to be involved in many projects and talk to many people simply as a necessity to create tools.
Although I don't know you and I'm certainly a "nobody" in the blogosphere, because of tight connections through blogrolls and such, I could probably find a connection in 2-3 degrees of separation. Actually, it's probably more like a few dozen connections in a very small radius.
The way the blogosphere works (lots of connections going in every direction) means that for information to pass around a great deal of the 'sphere requires only a few short passes to the big nodes (like Mark Pilgrim, Boing Boing etc) and then back out to the masses. Very fast, very efficient, very clever.
Now, if only we could utilise these connections in a more direct manner...
GaryF: FYI: I work for a large company. I work on interoperability with other companies both large and small. I achieve significant cooperation despite the backdrop of competitive differences.
Good! Glad to hear that at least some people find it easy to get interoperability on projects. Any time I've seen companies "work" together on projects (not actually involved myself), they've generally been trying to skew the project to their own needs or, even worse, trying to screw the other company over.
In a sense, I earn most of my wages by acting as an Information Gateway, both between Apache developers and users and English speaking communities and Spanish ones. So, please consider in your analysis non-transparent proxies like myself (either translating languages, training, or adapting information to narrower/vertical scopes).
I have also seen a big increase in the number of neighbors I'm switching between through Open Source and blogging :-) (though comparing with Sam would be like comparing a small ADSL router with a big Fiber switch)