[SPG_Active_Members] Long-term Digital Preservation: your technical questions of Sept. 1

[Al Kossow]

On 9/7/09 7:03 PM, "Tim Shoppa" <shoppa at trailing-edge.com> wrote:


> Maybe the lesson is to not write end-all-and-be-all
> tools but to have simple tools; this implies (I believe) simple
> formats for disk/tape images.
> 

That is the philosophy I have been following while imaging media, along with
trying to find documentation on every recording and file format that I can.

Doing so paid off when I was given several hundred TENEX tapes to archive.
The format was documented in some papers in the SRI document collection we
have. I also just found out someone at Stanford had to recover some SUMEX
backups about five years ago, and had to solve the same problem (recovering
bsys backups).

An alternative approach was the MIT Time Capsule File System, where they
tried to do as much up-front decoding of the tape images as they could,
assuming the knowledge on how to interpret the dump tapes would be lost.

All of the AI Lab's archive tapes were recently recovered, and imaged in
'.tap' (low-level image) format. I haven't heard if they plan on creating
TCFS images from the new data (several thousand tape images).