Skribent: Lars Noodén Dato: Til: dng Emne: Re: [DNG] a how to question (project(s) related)
On 8/17/21 10:39 AM, Dr. Nikolaus Klepp wrote:
[snip] > I've lost all my earier works on amiga and plus4 due to bitrotting -
> I sold the hardware when I realised everyting on floppy dodn't even
> last some years.
The life span of various storage media under various storage conditions
was (and is) well-documented but widely ignored. As a student and,
later, researcher in libraries and archives in the 1990s 'we' tried hard
to inform people of what the dangers were and of viable preservation
strategies. Now as then ways out mostly center around migration from
one storage medium to another. Filesystems that detect flipped or lost
bits, like OpenZFS or BtrFS, help a little there.
Migration from one data format to another is riskier, more difficult,
and more error prone. So for that the likely way out would be
emulators, possibly layers of system emulators in order to be able to
run the newest remaining software still capable of accessing and
rendering the data. FWIW you can emulate an Amiga easily now.
I have been migrating some random files from the mid 1980s but nothing
of importance. I still wrote college essays directly on paper then and
those papers have gotten lost. I do regret erasing the 'extra' backups
I had of a key WWW project from 1994. That whole project could have
been saved with some cost but little effort. The size it was then is
now considered insignificant these days. Too many other files since
then have been lost to flipped bits.
Regarding the original question from o1bigtenor about drowning in one's
own documents, if the file formats support embedded metadata, and there
is an indexing program which can process that metata, then that is one
way. So a mixture of HTML, ODF, and PDF, each with embedded
keywords/subject headings, title, date, and maybe a description can be
indexed by Recoll or several other tools by the available metadata.
Another way would be to expand on the example in the original message
and try to group the apparent locations of each item. Zim can do that
in an annotated manner by linking to the local documents. So it would
even work for turning the document collection into a zettelkasten.