The Reader Response Issue: FileSlinger™ Backup Reminder 01-12-07

Last week’s backup reminder prompted some reader responses, which I thought I’d share with you this week.

The first was from Loyal Reader Mike Van Horn:

Great philosophical discussion! What medium do you use to record information to last for what period of time?

Day, week, month, year
10 years,
100
1,000
10,000
1 million years

We send data etched on gold plates along with a space ship that is headed on a long slow journey beyond our own solar system. If found, may be a million years from now.

We read DNA preserved in tissues . . .
10 thousand year old Neanderthal
100 million year old dinosaur bone

We’ve got our Rosetta Stone (3k years?), Dead Sea Scrolls (2k + ), inscriptions inside the Pyramids (5k?)

In my closet, I have my handwritten notebooks going back into the early 1960s, easily and instantly readable.

But I also have 128k disks from mid-1980s that I have no way of reading.

The only reasonable way to back things up for extended periods of time:

• Durable media
• Varied forms
• Dispersed
• Hidden
• Remember where you hid them and how to use them

My response was as follows:

Gold is pretty durable, as long as humans can’t get their hands on it. Ditto stone and so on. But unless it’s kept away from people, the likelihood of re-use is a risk with things like gold, bronze, and marble. I used to be a classicist, and I know very well how many ancient monuments were melted down (if bronze) or hauled off (the facing on the Colosseum) or put into lime kilns (a lot of marble statuary). And then there are the ancient manuscripts which might still have been readable if they hadn’t been deliberately or ignorantly destroyed. Even the conditions which have preserved papyri for millennia have been pretty hit-or-miss, and the written work that has survived is a tiny fraction of all that was created and even of all that was highly valued in its own time. Fire, flood, or mudslides could make those notebooks of yours difficult to read or destroy them altogether.

Then there’s the issue of ease of creation. It’s hard work carving out stone blocks, and it’s expensive buying gold plates to inscribe. In Egypt, literacy was restricted to a very few people. Part of our problem with storing data now is that it’s so much easier to create data.

But the copy-it-over method does work. Most of the ancient works that remain to us today exist in medieval manuscript form, copied over and over and over. We have scraps of Greek literature found on papyrus, but those, too, postdate the works’ composition dates by decades or centuries. If people hadn’t set out to make new copies as the old ones wore out, we’d hardly have anything.

Then the Ur-Guru got in on the game. Actually, he’d made one remark already about the ability of the Egyptians to preserve data, so I thought he’d want to read Mike’s comment and my response to it. And this is what he had to say:

Interesting discussion, indeed. 🙂

Gold, stone, granite, magnetic pieces of circular discs, it all comes down to a decision where cost and effort are directly proportional to the value of the content.

If there is 400GB worth of valuable data to backup it’s easy to copy it all to an external HD of the same size. Quick and efficient but less reliable than having the 400GB on good quality DVD’s. The DVD’s would almost certainly last a little longer than the HD. But the cost of the HD would be around $200 and the cost of 100 DVD’s of good quality at around $1 each would be doulbe that. Add to that the time needed to play disc-jockey with 100 DVD’s and that’s a higher cost at more effort, depending on how valuable the content is. Doing both makes even more sense to get the best of both worlds, but again ups the cost.

How long do we want our content to be saved and recorded for history is another aspect. Does my source code matter 1000 years from now? Or even 100, or realistically speaking even 10 to 20 years from now. Very unlikely. I’d like my pictures to be recorded for the next eons but how realistic is that, to expect anyone to be interested in them so much further in time. On the other hand the Egyptians and pre-egyptian civilizations that have built some of the huge structures did not only build them to hold inscriptions of information but the structures and layout of the structures themselves are part of the message they have wanted to preserve for a long time, and successfully so. The value easily outweighs any of our digital pictures, source code, data so the effort was definitely proportional to the importance or value of the content there.

The biggest problem, though, is humans. As history has proven, all forms of records are evidently not appreciated or valued by the many generations that came after them. Robbing and pillaging, burning and destroying. Aliens that might at some point in the future find our gold plates might end up using them as coasters! 🙂

And I would add to that the observation that while it’s easy enough to read handwritten notebooks from 40 years ago, there are actually very few people who can read ancient inscriptions and manuscripts. In general, the manuscripts are harder than the inscriptions. The Romans, at least, liked to make their inscriptions in neat, consistent capital letters. The Greeks on the whole did the same.

Those inscriptions bear very little resemblance to the scripts used by scribes on papyri, to the point where someone who can easily read the Oxford Classical Text version of an ancient work might have no better luck reading the oldest manuscript of the same work than someone who had never studied Greek or Latin. (Trust me: I am speaking from the experience of my former life, and it was a shock to me.) I can only assume this is just as true for Hebrew and Egyptian, and I know that cuneiform changed drastically over time and depending on the culture using it. (The cuneiform script was used to write many different languages.)

The fact that something exists, whether it’s an inscribed stone or a 128K floppy disk, doesn’t mean people can read it.

Fortunately, as Stefan said, we don’t actually need most of the data that we back up to either last for millennia or be usable that far in the future. We just need to preserve it long enough to guard against tax audits.