Most of my weblog activity – both reading and writing – has historically had a technical focus. Recently, though, I’ve been trying to explore more non-technical weblogs, and tonight I decided to use Steven Den Beste’s USS Clueless as a starting point. Yeah, I know Steven’s a hawkish libertarian/screwyouitarian twit sometimes, but he’s the kind of twit who provokes thought. Anyway…it didn’t take me long, as I was working through his “blogroll” to find the Blogosphere Ecosystem. Maybe everyone else knew about it before I did, but just in case I’m not the last I’ll describe it. It basically scans the front pages for a list of registered weblog sites, looking for links to other sites on the list. Then it counts up the links to each site, sorts the results, and displays rankings.

Sound familiar? It’s a lot like Google’s PageRank, or the sort of “distributed reputation system” probably best exemplified by Advogato. One thing I naturally wonder about, having made those comparisons, is how resistant the ecosystem’s tally is to deliberate tampering. If I and a bunch of friends all register our sites, and then fill up our front pages with links to each other, will we get ranked near the top?

I also wonder how much the results are dominated by the linkage to/from the highly-rated “warblogs” that the author started with. If you remove the top twenty or so from the equation, how much do the standings among the rest change? I imagine you could find all sorts of separate clusters if you really wanted to. Then again, maybe not. The tech-blog community hardly seems visible at all in the ecosystem’s lists. Similarly, the gay-blog community seems to be one of the most cohesive on the net and large enough that I run into it frequently even though I have little interest in gay issues, but they don’t seem represented either. Maybe we need multiple ecosystems. Maybe, even better, we need some automatic way to take a big unwieldy graph of links between sites and break it down naturally into separate clusters.