If someone is going to make the effort to decode the information from
the original computer media into the archive standard then if the data
is 10 or 100 times larger than the original should not be an issue. The
same is true for costs. I have seen too many systems where there was a
perceived decision that it would be too costly to "do it right".
The idea is to take a copy of a program or data and convert it/archive
it once and know that it is then saved.
Think of the following;
Stone once was expensive so clay tablets were used.
Papyrus was expensive so wax tablets were used
At one time paper was too expensive so it was not used chalk on a board
was used instead.
Later pulp paper was cheaper and acid free paper was expensive.
Punch cards once were cheap
Mark Sense forms
Magnetic tape
Magnetic disk
Optical disk
CD-ROM
DVD
The "best" real long-term media that I can think of is "eye ball"
readable. Think Rosetta stone.
Do you want to use Big-Indian, Little-Indian, Intel, Motorola, Cyrillic,
cuneiform, hieroglyphs, binary, hex, octal RAD50, left-to-right,
right-to-left, Chinese, Japanese, English, Pig Latin, Latin, Spanish,
Old English, Old Latin, Norse, Mayan or Esperanto?
I have heard that the Mormon Church is using iridium disks inscribed
with an ion writer, it may take a magnifying glass but it's readable.
It's supposed to last 1,000 years.
This may be slightly sacrilegious. We will then need a "priesthood" of
fanatical zealots to protect the archived data for posterity. I
nominate Sellam to be the patron saint of data retention. Tony can be
the patron saint of hardware restoration.
I'll be friar Mike hunched over a lectern transcribing data with an ion
writer.
Mike