In article <1993Jan22.175120.23156 at gserv1.dl.ac.uk>, ajt at rri.sari.ac.uk
(Tony Travis) wrote:
>> In article <93122162041.MIN-LDGBa29809.bionet-news at uk.ac.daresbury> you wrote:
> : [...lab notebook stuff]
> : 3. No one has mentioned published encrypted summary data. Somewhere
> : I read about (but have not personally seen) services which take any
> : data file and produce from a large file a small "signature" which
> : is published in the classified section of a microfilmed newspaper.
> : Experts will testify that it is essentially impossible to produce
> : a file which matches the signature and is not the original.
>> Am I totally misunderstanding what you say, or are you just describing
> the function of CRC's ?? The combination a file's length in bytes and
> its 32-bit CRC (cyclic redundancy check) is virtually unique and serves
> to validate many software packages. This, in essence, is no different
> to validating experimental records as original.
The purpose of the newspaper classified ads containing these signature
numbers is to be able to prove in court that a file containing the data
existed prior to the date of publication of the newspaper. A simple
CRC would not be adequate because alterations of the data would be
feasible: if you want to change one number somewhere in your file,
you just need to find out a compensating change somewhere else that
gives the same CRC. The encrypted signature function makes this
impossible. This answers the objection to computerized lab notebooks
which asserts that they can be altered after the fact undetectably.
>> : What is needed to make computerized lab notebooks feasible
> : is a similar service on the internet (maybe BioSci could offer it)
> : where "signature" files could be publically posted and permantly
> : archived in such a way that they could not be modified without
> : the cooperation of many people (and no one could be certain that
> : someone else didn't keep another copy somewhere). E.G. create
> : a newsgroup for these files, and let anyone who wants to archive
> : it.
>> What is needed to make data accessible is ISDN!!
I was addressing the validation problem, not the accessibility problem.
But ISDN (or better, fiber optics carrying cable TV and internet service
to each home) would help.
Edward J. Huff huff at mcclb0.med.nyu.edu (212)998-8465
Keck Laboratory for Biomolecular Imaging
NYU Chemistry Deptartment, 31 Washington Place, New York NY 10003
Posted from the new Mac NewsWatcher 1.3d5