Where is the Internet’s memory, the history of our time?

January 20, 2015 8:17 AM

“Every time a light blinks, someone is uploading or downloading,” Kahle explains. Six hundred thousand people use the Wayback Machine every day, conducting two thousand searches a second. “You can see it.” He smiles as he watches. “They’re glowing books!” He waves his arms. “They glow when they’re being read!”

The Cobweb: Jill Lepore on whether the internet can be archieved, the Wayback Machine, the downing of Malaysian Airlines Flight MH17, the complications of attempting to put a time dimension on a two dimensional medium and the almost destruction of the footnote. Featuring a cameo by MeFi's favourite archivist, Jason Scott.
posted by MartinWisse (7 comments total)

Cool. Has anyone tried this?

Herbert Van de Sompel, a Belgian computer scientist who works at the Los Alamos National Laboratory, is trying to reweave the fabric of the Web. It’s not possible to go back in time and rewrite the HTTP protocol, but Van de Sompel’s work involves adding to it. He and Michael Nelson are part of the team behind Memento, a protocol that you can use on Google Chrome as a Web extension, so that you can navigate from site to site, and from time to time. He told me, “Memento allows you to say, ‘I don’t want to see this link where it points me to today; I want to see it around the time that this page was written, for example.’ ” It searches not only the Wayback Machine but also every major public Web archive in the world, to find the page closest in time to the time you’d like to travel to. (“A world with one archive is a really bad idea,” Van de Sompel points out. “You need redundance.”) This month, the Memento group is launching a Web portal called Time Travel. Eventually, if Memento and projects like it work, the Web will have a time dimension, a way to get from now to then, effortlessly, a fourth dimension. And then the past will be inescapable, which is as terrifying as it is interesting.

posted by stbalbach at 8:51 AM on January 20, 2015 [2 favorites]

(Paywalled, if you've used up your monthly quota, but the usual workarounds work.)
posted by dhartung at 8:51 AM on January 20, 2015

Coral Cache version, for irony points.

Holy crap I can't believe Coral Cache still works.
posted by Kadin2048 at 9:03 AM on January 20, 2015 [4 favorites]

This is a case of the highly positive, true believer. No take backs!
posted by Oyéah at 10:04 AM on January 20, 2015

whether the internet can be archieved

That typo is quite an archievement .
posted by w0mbat at 1:23 PM on January 20, 2015 [2 favorites]

Cool. Has anyone tried this?

I work on Perma, another service mentioned in the article. I saw a beta of Time Travel recently -- it's very neat. It looks just like a Google search results page, except you put in a URL and each result is that page from a different archive on a different date.

The Memento Chrome plugin is available now, though, and really demonstrates the potential of this kind of tech. It adds a time dimension to links. For example, you can go to a New York Times article, right-click on a link inside the article, and it will offer to follow the link as of the date the article was published.

This addresses the same problem we've talked about in Metatalk, where the links in old posts are mostly broken. With the Memento plugin, and with a standard datePublished meta tag in the Mefi header, you could right-click and go directly to the closest available snapshot in any Memento-compliant archive.

This stuff is still a little clunky right now, but it's a great demo of what's possible. I think as the web gets older we'll come to see the time dimension of links as indispensable.
posted by jhc at 6:47 PM on January 20, 2015 [1 favorite]

Interesting article. It's a complex problem space, it's worth noting that all these efforts (archive.org, europeana/dpla, memento, perma.cc, etc.) are not so much variations on a theme, as fundamentally different approaches. The perma.cc approach is eminently sensible I think, at least for that specific purpose. There continues to be a lot of interesting technical and philosophical discussion in this community on what a document is, what an archive is, what a collection is, and also on how to describe and catalog these.
posted by carter at 7:54 AM on January 31, 2015