The GeoCities deletion

What the volunteers got

Archive Team crawled it for months and released the result as a torrent anyone could seed: about 640 GB compressed, roughly 900 GB unpacked. Asked what share of GeoCities that was, Jason Scott answered:

How much? We'll never know.

They phoned Yahoo! to ask how big the service was and were refused. Their guess was around 10 TB against the 23 million pages Yahoo's own Site Explorer claimed, with the caveat that they could be wildly off in any direction.

You will see "38 million pages" quoted everywhere. It traces to Wikipedia, which the Internet Archive quoted into its own collection description, which people now cite as though the Archive had counted. Nobody counted. The only people who could have said were the landlord, and they declined.

Field marks

• Three levels deep, and the last one is a number. geocities.com/area51/1013. Nobody has shipped an address like that since.
• The street is the subject. /area51/stargate, /area51/dungeon. The URL was a heading, which is how discovery worked before search.
• The map has typos in it. dungeom beside dungeon, labrinth beside labyrinth. Both pairs are on the real index. Somebody typed it by hand.
• The closure notice is a cross-sell. The most telling artifact in the unit is what the landlord thought was a replacement.

What it solved, and what took it

You had an address, and the address was a place, and your neighbors were people who cared about the same thing you did. Platforms replaced the address with an account, and an account can be closed.

Back to the directory.