memory

Last night we came home from Vacation Church School to discover a family tragedy: our iMac's hard drive is expiring, fast. The computer has been declining noticeably in recent weeks, and according to our in-house Apple geek, last night was time to call for the iPriest to administer last rites.

I just had a conversation with BQO webmaster David Dylan Thomas, who asked when we bought the computer. In 2005, I told him. Yep, he said, it's time to upgrade. Julie and I decided we'd get a Macbook Pro -- having a laptop for our main home computer is more useful and convenient for us than a desktop model -- so now the only question is whether or not to get a 13" model or the more expensive 15" model. We really don't need all the extra processing speed in the 15" model, but I'd feel more comfortable having that extra space and speed built in, because I don't know three or four years from now what our needs will be. If you have advice, let's hear it, because I'm going to drop a wad of cash on the Apple store real soon. Gloom!

The sad thing is, Dave continued, is that these days, they build computers with the thought of obsolescence in three years. We are so forward-moving in this culture, he said, that nobody thinks about backwards compatibility. True: just the other day, I was in the Apple store looking to buy an upgrade to Snow Leopard, and learned that it's impossible to upgrade my iMac from the Tiger operating system to Snow Leopard. It just can't be done.

Anyway, check out Dave's comments on Jeffrey Rosen's much-discussed Times piece about how the Internet never forgets. It may be theoretically possible to store everything about our online lives in perpetuity, but this isn't practically true. Excerpt:

As a web manager, I've overseen the overhaul of many a content management system, and there's always a compatibility issue which forces editors and technology teams to ask the same question. How much? How much will it cost (in time and money) to convert how much information? Do we really want to bother reformatting 400 news stories that were published in 2000 to a whole new format on the off chance that someone will search for them? The answer is almost always no. And that's just 10 years.

We assume that formats like .jpg (that picture of you doing a kegstand) or .mp3 (that ill-advised phone message you left at 3am) or — I won't even pick a video format since they change every week — will be here forever because they've been around as long as we can remember consumer-friendly digital information. But the odds that your Facebook page will still be here in ten years — or will be readable in ten years — while not terrible, probably aren't as good as you think.

People have been talking about this problem for a long time now. And they still are. The problem we face may not be that the web never forgets, but that it is better at forgetting that any technology we've ever used. 

Read the whole thing. I have a box at home of VHS tapes, home movies from the 1990s, and a few cassettes with personal recordings that mean a lot to me. I am realizing now that I have nothing in my household that can read the information on those tapes. I could go to the store now and buy a VCR and a cassette player, but I'd have to look hard for either. Ten years from now, will the technology to access the information stored on VHS and audiocassettes even exist? What happens to those "memories," then? Think about that before you build your e-book library.

Jeffrey Rosen has an engaging piece in the Times about privacy and the web that touches on issues of forgiveness and reputation and how the Internet has basically screwed that up for all of us; the upshot being that because your Facebook profile never really goes away, your sins are plastered on the world's largest wall for all to see forever.

Here's the thing. They're probably not. Forever, that is.

Digital archivists would love if Rosen's assumptions about online inertia could be taken at face value — anything that goes on the web, stays on the web unless acted on deliberately by a person or built-in self-destruct. But that's not the only way — or the primary way — things disappear from the web. Anyone who's looked for a link to a really cool article only to find that it just isn't there anymore knows that digital information is all too mortal. Incompetence and neglect are just as likely to kill a piece of web ephemera as a cease and desist order.

But some blogger who refuses to use permalinks or Yahoo news letting their old urls die is only the beginning of what keeps digital archivists up at night. If you want to know what else does, scour your home office and see if you can find a floppy disk. Found one? Okay, tell me what's on it.

See, the web is good at a lot of things, but legacy is not one of them. Technology in the past three or so decades has been a story of improving standards and formats but backwards compatibility? Not so much. 

This is true for more than just physical formats. File formats are what will be the death of history on the web. Take the Flash vs. HTML5 debate. Let's say HTML5 wins. How many sites that render in Flash will convert all of their data to render in HTML5 and not just the most recent two years worth? Or six months?

As a web manager, I've overseen the overhaul of many a content management system, and there's always a compatibility issue which forces editors and technology teams to ask the same question. How much? How much will it cost (in time and money) to convert how much information? Do we really want to bother reformatting 400 news stories that were published in 2000 to a whole new format on the off chance that someone will search for them? The answer is almost always no. And that's just 10 years.

We assume that formats like .jpg (that picture of you doing a kegstand) or .mp3 (that ill-advised phone message you left at 3am) or — I won't even pick a video format since they change every week — will be here forever because they've been around as long as we can remember consumer-friendly digital information. But the odds that your Facebook page will still be here in ten years — or will be readable in ten years — while not terrible, probably aren't as good as you think.

People have been talking about this problem for a long time now. And they still are. The problem we face may not be that the web never forgets, but that it is better at forgetting that any technology we've ever used.