Archiving information, was Re: ADM-3A question

Seth J. Morabito web at
Fri Aug 16 10:31:42 CDT 2019

Paul Koning via cctalk writes:

> Anything worth having around deserves backup.  Which makes me wonder
> -- how is Wikipedia backed up?  I guess it has a fork, which isn't
> quite the same thing.  I know Bitsavers is replicated in a number of
> places.  And one argument in favor of GIT is that every workspace is a
> full backup of the original, history and all.
> One should worry for smaller scale efforts, though.

This is a problem I think about a lot.

In the early 2000s I worked on the LOCKSS program at Stanford
University. LOCKSS stands for "Lots Of Copies Keep Stuff Safe", and is a
distributed network of servers that replicate backup copies of
electronic academic journals. It stemmed from a research project that
looked at how to design an attack resistent peer-to-peer digital
archival network.  Each node in the network keeps a copy of the original
journal content, does a cryptographic hash of each resource (HTML page,
image, PDF, etc.), and participates in a steady stream of polls with all
the other nodes where they vote on the hashes. If a minority of nodes
loses a poll, their content is assumed to be damaged, missing, or bad,
and they replicate the content from the winners of the poll.

It's designed as a "Dark" archive, meaning the data is there, but nobody
tries to access it unless the original web content disappears. Then, the
servers act as transparent web proxies, so when you hit the original URL
or URI, they serve up the content that's now missing from the real
public Internet.

It's a neat idea. It's also open source, and unencumbered with
patents. I've always thought a similar model could be used to archive
and replicate just about anything, but it's just one of those things
that nobody's ever gotten around to doing.

> 	paul


  Seth Morabito
  Poulsbo, WA, USA
  web at

More information about the cctalk mailing list