Former purveyor of discarded paradigms, at your service.

Designer Richard Vijgen has posted a lovely and ambitious visualization of a data set that has special personal meaning for me: 650 gigabytes which represent tens of millions of homepages hosted by GeoCities before it was shut down by Yahoo! in 2009. From 1994-1999, I worked at Tripod.com, leading competitor to GeoCities. When our larger, more successful competitor shut down in 2009, some of the people who’d been involved with founding Tripod traded emails, congratulating ourselves on the fact that our site still exists and still hosts homepages, even though ownership of the company has changed hands several times. But that celebration was hollow – we may built one of the ancestors of today’s participatory media platforms, but our glory days are long past.

The Deleted City from deletedcity on Vimeo.

Vijgen describes his project – which visualizes the filesystem of the GeoCities archive as a vast city – as a form of “digital archeology”. It’s an interesting term to use – while archeology is the study of civilizations through the study of their artifacts, it’s often associated with the study of long-dead civilizations. GeoCities is an abandoned city as much as a dead one. Yahoo! shut it down after concluding that there wasn’t enough traffic to the millions of homepages to justify selling ad inventory on them or continuing to pay the server upkeep and maintenance costs. (My guess, based on helping run a similar company – removing copyrighted content and dealing with abuse complaints was likely another major cost for Yahoo!)

People stopped caring for their pages and moved on. The pages persisted, as digital things do, unchanged from the last time they were tended to. (At a conference in Chiba in 2005, a scholar told me that a term had emerged in Japanese to refer to abandoned blogs. The term was similar to the word for “river stone”, implying something solid, unmoving and mute.) As an archeology project, visualizing Geocities is more a study of Centralia than Chaco Canyon: we know who left and why. If it’s worth studying the structures left behind, it’s not to solve a mystery. It’s to understand the shift that’s taken place.

“Ten years later in 2009, as other metaphors of the internet (such as the social network) had taken over and the netizens had moved on to Myspace and Facebook, Geocities was shut down and deleted.” The shift in paradigm that Vijgen describes has two dimensions. He’s talking about a shift from a geographic metaphor that Geocities was nominally organized around, where people with similar interests located in the same “neighborhoods”, lovers of rural life in “Heartland”, gays and lesbians in “South Beach”, etc. (We studied GeoCities neighborhoods at length at Tripod and decided that affiliation was little more than nominal – there was no zoning that prevented off-topic pages from appearing in the “wrong” neighborhoods, and very little ability to predict what your “neighbors” were interested in.)

The more important shift in metaphor was from pages to streams. In the mid-1990s, we understood the web in terms of pages. Some pages were meant to be permanent, others changing, others completely ephemeral. Blogs updated the paradigm somewhat – they were pages we expected would change, daily, perhaps weekly. But they were pages, with permanency and permalinks. And you controlled what went on them, even if you permitted comments on your blog. Conversations took place between spaces – I link to you, you link to me. In the age of Twitter and Facebook, pages feel too permanent, too fixed. You produce a stream of updates which flow past your friends. If they follow you closely, they might hang on your every update – more likely, they dip their feet into the stream now and again, seeing what you’re up to, chiming in with a comment or an upvote.

It would be a mistake to visualize these interactions as buildings composing a city. They’re rivers, distinguishable by path and magnitude, but shifting and ever-changing. You can never step into the same lifestream twice.

We weren’t total idiots in the mid-90s. We knew that pages weren’t the right metaphor, weren’t going to be state of the art forever. We could see communities emerging in systems like Webrings, which tied disparate pages together into a loose aggregation. (When Charley Lanusse, the creator of Webrings, wisely rejected our overtures to buy his company, Tripod built “pods”, our version of the same basic tech. Charley later sold his company to GeoCities…) But knowing you’re a purveyor of a dying paradigm isn’t the same thing as knowing what the next big thing will be, or being able to build it.

Paradigms can shift quickly on the Internet. It’s hard to imagine anyone unseating Facebook, especially given the limited traction Google+ has achieved despite valiant efforts. But not everyone thought Yahoo! was insane when they paid billions for GeoCities, or when Murdoch bought MySpace. That sense that Facebook isn’t quite what we want, blurs boundaries between communities we want to keep separate, isn’t respectful enough of our privacy or our ownership? To me, that’s an indication that the paradigm is not quite right and ready to shift. To what?

I have no idea. But that sense that the ground is moving would make me reluctant to invest too much time, money or energy in Facebook. (Then again, it’s worth remembering that Tripod made lots less money than GeoCities, in part because we got out too soon. We weren’t making money selling ads on homepages, couldn’t see that changing and figured we’d better get out when we could. GeoCities waited longer and sold for much more. So did much smaller, less successful companies like TheGlobe.com. Knowing the ground is shifting isn’t necessarily good for your fiscal health.)

I chaired a panel at Microsoft Research the other day on privacy, where three very smart researchers offered their takes on what we do and don’t understand about online privacy. (Chris Conley of the Northern California ACLU summed it up nicely: “We say we care about privacy. But that’s not how we behave.”) In the conversation over drinks that followed, we got the inevitable question about big data: is there any escape from the masses of data that marketers are collecting about our every move?

I’m starting to feel like a contrarian on this question. Yes, my browser is riddled with cookies, and yes, ads for treadmills now track me across the web because I was browsing exercise equipment on Amazon the other day. But I’ve been less worried about this since Doc Searls pointed me to Rapleaf, a company that claims it “wants every person to be able to have a meaningful, personalized online experience.” It achieves this meaningful experience by selling your personal data to advertisers so they can better target ads to you. And they’ll let you retrieve and review the info they have about you. Despite the reams of personal information I’ve posted on this blog and social media sites across the web, Rapleaf believes I’m an unmarried, childless, underemployed likely Republican voter.

Should I worry about companies storing my data if they seem incapable of drawing correct inferences about me? Is the agglomeration of personal data more worrisome than the very real tendency of old bits to decay and disappear? GeoCities survives only because the Archive Team sprang into action to save it. Other sites and communities expire every day. Are we being wise, or paranoid, when we assume that our movements will be tracked forever in massive databases, while our utterances and creations have a tendency to expire and disappear?

Seeing GeoCities as an archive and an art project makes me feel old. In the dot.com days, people talked about internet years as if they were dog years – a company around for 4 years was a stable mature adult of 28, and an eighteen-month old startup might be worth taking seriously as if it were entering its early teens. This blogger makes an interesting case for 4.7 internet years to a calendar year – by his math, I’m over 100, fully entitled to shoo these twittering youngsters off my lawn with my cane and to wander GeoCities long-dead neighborhoods, remembering when this used to be a cool and trendy neighborhood.

Does history slow once the Internet is no longer the province of alpha geeks who decamp for greener pastures as soon as they’re mapped? Once we’re all on Facebook, will we never leave?

Take it from an old man who’s watched tumbleweed roll through the streets of a deleted city. Everything changes. Nothing lasts. That’s a good thing.