Re: [OpenAFS] Accuracy of timestamps in AFS files.

[Chuck Boeheim]

I had a similar experience, and ended up trawling through our tape vault, and finding backup tapes of the first web server in North America from January 1989. I had to document where and how they had been kept, and then take them to a local large manufacturer of mainframe computers where they still had the hardware and OS to read the tapes. I had to document every step, and answer a day and half of detailed questions about how the files were handled.

Hope this helps you.
-Chuck Boeheim, SLAC


On Apr 19, 2008, at 8:31 AM, Finke, Jon E wrote:

A useful source for you might be to find a backup tape from before the intellectual property case came to light. It would be hard for the other party to argue that you would go back and change the dates on backup volumes from before you knew there was an issue. Having multiple backup tapes showing that the last changed dates on the files had been consistent and unchanging can be used to establish that those dates are credible evidence. You should consult with your counsel about how to establish the proper chain of custody of the evidence before you touch any tapes. Our evidence of prior art helped to have the patent thrown out, which is why the world doesn't have to pay a certain individual royalties every time a web server does a backend search of a database. We were able to show that someone at SLAC did it first and freely published the technique.

I will be giving a deposition involving some intellectual property issues in the near future, and some of my work is being considered as prior art. One of the ways we are determining “when” I wrote things will be based on the last changed dates of the files. Our AFS cell dates back to the early 90s (which is when I got my AFS admin certification.)

smime.p7s
Description: S/MIME cryptographic signature