DailyDirt: Keeping Information For A Really, Really, Really Long Time
from the urls-we-dig-up dept
The problem of storing digital data usually involves transferring data from an older format to a newer one — with the hopes that the newer one won’t be replaced as quickly as the older format it just replaced. Maybe some archivists out there like to go through this periodic technology shift and verify that the data we’ve stored is still readable, but wouldn’t it be nice if there was a “store it and forget it” format?
- A single gram of DNA can store up to 455 exabytes of data, and roughly four grams of DNA could store all of the world’s collected information. It’s not cheap to store arbitrary data on DNA yet, but the costs are coming down. [url]
- Stored data on DNA encased in glass at freezing temperatures could survive for a couple million years. The DNA of a woolly mammoth in Siberia was preserved for 40,000 in a tundra, so it’s not unthinkable that we might want to keep a few backups of our data on ice. [url]
- Making large DNA crystals with precisely-controlled structures is something that can be done with a “DNA-brick self-assembly” method. Being able to do this could help fabricate all kinds of nanoscale designs and create a manufacturing process for complex, artificial DNA structures. [url]
If you’d like to read more awesome and interesting stuff, check out this unrelated (but not entirely random!) Techdirt post via StumbleUpon.
Filed Under: archive, backup, data, dna, nanotechnology, self-assembly, storage
Comments on “DailyDirt: Keeping Information For A Really, Really, Really Long Time”
DNA is probably a good storage medium bc it’s not proprietary and if people are still alive and technologically adept in the future, they’ll at least know something about how DNA can be used to store data
“wouldn’t it be nice if there was a “store it and forget it” format?”
The CD-R originally claimed to be exactly that “store it and forget it” archive medium — at least if we believed the promised claims from so-called “experts”. My own personal experience has been that CD-Rs (at least the early ones) could not be counted on lasting for more than maybe 5 years or so, and after 15-20 years a high percentage were unreadable. And it wasn’t just the ability to read those microscopic dots: some CDs even had the dye layer visible separated around the edges. The original claims of 100 or 200 year life spans for ‘burned’ CDs turned out to be pure bullshit.
Although DNA can be extracted from the (non-fossilized) remains of ancient animals, it suffers the same problem as magnetic and optical storage media: degradation that leaves many “holes” in the archive. So there’s no chance of cloning a woolly mammoth … yet. Maybe sometime in the distant future someone will develop a way to combine intact (partial) strands of DNA from many different cells into one single complete DNA record, thereby making cloning possible.
Re: Re:
“The CD-R originally claimed to be exactly that “store it and forget it” archive medium — at least if we believed the promised claims from so-called “experts”.”
I think you’re confusing “the manufacturers” with “experts”. Manufacturers made (as they always do) ridiculous claims of the longevity of CD-Rs. However, the experts that weren’t getting paid by the manufacturers always said the same thing about CD-Rs: they have a limited lifespan. In office conditions, this would be about 5 years. If you take care to store them in ideal conditions (no humidity, no light, and frozen) they can last more than 20 years.
Also, in terms of longevity, there’s a real difference between cheap CD-Rs that you can get for $15 per 100 and high quality CD-Rs.
This looks interesting, but the thing I’m a bit worried about is, if it’s possible to create arbitrary DNA just by designing it on a computer, then it’s possible to create biological DNA just by designing it on a computer. Does anyone really want to see the next generation of malware uploaded to the Web be viruses of the bio-warfare variety?
Re: Re:
DNA made by a computer and brought to life? Done. https://en.wikipedia.org/wiki/Mycoplasma_laboratorium
“I think you’re confusing “the manufacturers” with “experts”. “
The “so-called experts” would be the tag-team combination of the manufacturers and the lapdog mainstream media (who generally don’t want to upset their advertisers). I got into a lot of arguments with people about these kind of claims back in the day, insisting that any claimed shelf life on a new product was theoretical at best, and that there was simply no way to predict how long something would actually last until proven in real-world conditions, and that anyway, most products based on new technologies often have a much shorter life than the manufacturers’ predictions.
I was making these arguments at the time there was a huge billion dollar class action settlement for homeowners who had installed polybutylene plastic pipe, which turned out to have a much shorter lifespan than people were led to believe. So I was always a bit leery of the claimed 100+ year lifespan for this other new technology, CD-Rs.
But what did I know when ‘everyone’ from CNN to the NY Times was drinking the Kool Aid!
“Also, in terms of longevity, there’s a real difference between cheap CD-Rs that you can get for $15 per 100 and high quality CD-Rs.”
I somewhat expected that, but oddly enough I had more trouble with the “name brand” CDs than the (easily scratched) cheap generic silver ones, all of which held up quite well. Anyway, using an ATIP reader showed that many of them were the exact same Ritek brand CDs sold under various different labels.
Re: Re:
“But what did I know when ‘everyone’ from CNN to the NY Times was drinking the Kool Aid!”
And they always do. It’s a mistake to think that technical claims in these sorts of media outlets are anything like accurate. All they’re doing is reading press releases, often verbatim.
“oddly enough I had more trouble with the “name brand” CDs than the (easily scratched) cheap generic silver ones, all of which held up quite well.”
That’s actually not odd at all. A “name brand” is not an indicator of quality (in fact, most “name brands” are the exact same discs as the generics, but with different labeling). You can’t tell quality from the brand or from the price. Your best bet is to search for “archive quality CD-Rs” and read reviews to select a high-quality blank that you can get through a local or favorite online retailer.
Re: Re: Re:
“most “name brands” are the exact same discs as the generics, but with different labeling)”
The ‘branded’ CDs being those with a thick, fairly scratch-resistent coat of paint on the topside of the disk, a protective layer which was completely lacking in the generic CDs, which needed to be handled very carefully — as I annoyingly discovered after writing my very first one. But the generics had one huge advantage: being able to write over the entire disk surface, not just the small white label area that the early branded CDs offered. The branded CDs were often (theoretically) much cheaper than generic, considering discount sales prices, and especially when counting the mail-in rebates that basically made them free — if you actually got any money mailed back (I never did — one rebate center even lost my paperwork twice).
It is the circle of life and we are back here. Storing data on the DNA is the most advanced and primitive form of data restoration.