Very good points.
You would think that it would not be terribly difficult to write software to transcribe an old digital medium to a new digital medium.
Certainly data compression algorithms work with little human interface. Algorithms ought to be able to be written to transfer data from one storage medium to another with little human interface and reduce the cost of updating archival storage.
You would think that it would not be terribly difficult to write software to transcribe an old digital medium to a new digital medium.
Certainly data compression algorithms work with little human interface.
-—<>-—<>-—<>-—<>-—<>-—
True to a point... the problem is that “holes” appear in the archived data. Magnetic glitches, formatting problems, re-formatting problems (a column or row was completely or partially shifted by inaccurate read/write of the previous archive record). Archeology (that is what we called it) is often needed, and that may require assistance of someone who was involved with the original record(s) to be reasonably confident of interpretation.
In addition, the ancient mag tape or paper tape machine may require maintenance, and those transistors or vacuum tubes are unavailable, so you have your electronics folks tied up attempting to reconstruct circuits that they have never contemplated. The mechanical problems with the old equipment are usually less stringent, but can be equally frustrating. Imagine how long it would take you to reconstruct the head of a 9-track tape machine... or the logic for verifying multiple checksums of blocks of information gleaned from a CDC (Control Data Corporations) written tape (Their computer information “words” were 60bytes, if I recall correctly... they certainly were not 64 bytes.) “Optical recognition” of data is definitely less than 100% accurate, and recognition of optical records that have been corrupted over time by assymmetric shrinkage of film or paper, as well as degradation of the media, exacerbates the problems.
These problems I’ve seen. I’ve personally dealt with some. I’ve got a few data tapes that I am not sure that even one machine exists in the world to read. Serious recovery of data archives is not at all trivial. Yes, a lot of automation can be used, but unless you are simply transcribing paper records to digital, embarking on this task quickly becomes very daunting.