In the Web archives

I’ve spent most of this week deep in the archival attic, researching the new book in old documents, digging through the dull roots of today’s Web, planted back in the 90s. It’s been strange and enlightening; I’ve found much interesting material.

One thing that becomes clear is that what we now think of as “the Bubble” was surprisingly brief. The Web actually experienced something of a downturn beginning in late ’97 and early ’98, and extending through the Long Term Capital meltdown later in ’98. It was only toward the end of ’98 that the bubble really began to inflate in a serious way. The High Bubble lasted till April 2000, when the market suddenly realized, like Wile E. Coyote poised in midair above a canyon, that it was standing on air.

So the era of high dotcom madness was really barely a moment: 18 months or so.

The other thing I’ve learned is how much more extensive the Internet Archive is than I’d realized. I’ve been using the archive heavily for days. I’ve picked up some pointers that, perhaps, others already know; I’ll share them anyway in case they prove helpful.

First of all, ignore all the error messages the Archive itself sends you, like “bad gateway” or “failed connection.” These are indicators of momentary failure; they don’t mean your page isn’t there. Try, try again; reload; eventually, you may get what you’re looking for. (On the other hand, error messages that are stored on target pages that represent the archive’s record of a snapshot of the web page itself — they’re real. They mean that the archive’s bot hit that error message and never recorded the page you’re seeking.)

Also: If the archive tells you that the earliest edition of a particular page it has is from, say, 1997, this doesn’t mean that the site’s content from previous years is gone forever. Iit’s true that you’ll probably never be able to recall, for instance, the Hotwired home page from 1995 — since it was constantly mutating, day by day with new content and year by year with new designs. But the material published on a site that lived at permalinked or semi-permalinked addresses can still often be dug up from Archive.org by poking your way carefully from the present into the past through the site’s own “back issues” or archives or “previously” links.

For instance, Web Review, the early GNN-backed web zine, vanished long before the Web Archive started up, along with most of GNN itself — a crib-death for one of the Web’s earliest original content ventures. Still, I was able to unearth my friend Andrew Leonard’s first piece (from Sept. 1995) for Web Review, all about “clickstream” measurement: Here it is.

We don’t have all of the early Web, but we have more of it than you might think!