Let’s get the numbers out of the way immediately before they’re misreported. Archive Team rescued, roughly, 20 percent of all the profiles on Friendster.  This took us many months and reflects the sheer mass of Friendster’s data, well into the 70-80 terabyte range, behind custom software, and which they summarily deleted from their user accounts.  We got maybe 10-15 terabytes of it, as best we could.

If you were firemen and going into a burning home, you’d grab the people, then the pets, and then you’d let the fucker burn because no fireman or person’s life is worth the risk. If you were the person in the house, however, you’d probably go after the other people and then grab family photos and maybe that sack of gold under the dresser and perhaps that Picasso you hang over the coat rack. In other words, you prioritize.

Now, instead of a fire, imagine it’s the Bagger 288 moving over the landscape towards an apartment complex with 112,000,000 apartments in it, and there’s this asshole shouting over a megaphone thanking everyone for living in their apartments for so long, and Archive Team is there with hacked-together go-karts flying down hallways and grabbing as much as they can with robot arms.

IT WAS EXACTLY LIKE THAT.

As a result, we prioritized ourselves – going after the first millions, then scattershot through the whole range of IDs, grabbing representative examples of profiles, up through to the deadly end, where we had Friendster accounts of our own (it took having profiles to acquire some information, and a couple brave Bothans who wrote custom scripts to yank out the un-yank-out-able). As a result, we got pretty unique directory structures, including blogs, photos, “shout-outs” and messages, all available to the general public. Oh yeah, Friendster killed tens of thousands of blogs, and millions of photos, and… great job, people.

This one was much more of a learning experience than Geocities, and the next time we’re faced with such a huge place going down in flames, we’ll be sure to use what we learned. Until then, we have these blocks of data.

We had one soul go through the entire userbase and groups of Friendster to produce this social graph/database of all the membership inter-relations, for the purposes of historical and academic study. Those aren’t too bad in size.

But the main data… that’s pretty big stuff.

But enough introduction: may I present the first million Friendster accounts.  There are still gaps as I put together the in-flooding collections from dozens of team members, which will take some time – the way the archive.org system works, I can inject the missing fils into the gaps as time goes on. Believe me, though, with 112gb of data, there’s plenty to check through as it is.

I know it might seem we should be proud of our work, but to be honest, I just consider it all with a blanket of sadness. It’s terrible this is happening – it’s awful that years of work are being destroyed in a ill-advised, greedy, misdirected moments. For years after this, people will go to check on their Friendster account, find it wiped, and find the whole thing has been sold down the river and is unable to help them get any of it back. With luck, Archive Team got it, but there’s no promise of that, and the form we have it in will be like finding out someone burned your place down (that metaphor again!) and Archive Team has a generic cardboard refugee box that we jammed some percentage of your stuff into and stuck into a massive shelf. That’s no replacement for the platform of expression and self and network of friends you had. None at all.

But it’s all we got.

How many more times?

Categorised as: Archive Team | computer history

Comments are disabled on this post