Business Columns & Blogs

As formats change, data migration strategies needed

Changing formats put our data at risk, and these days, data in the form of documents, photos, videos and texts are what we live by, which is why you can’t walk down the street without seeing people going the other way staring at their phones. Pokémon Go will be a transient fad but “augmented reality” gaming is a metaphor for how we overlay our digital selves on the real world. The key issue is retaining our data. The rising waters of this problem lap disconcertingly around our lives, and every time we think we’ve solved it, something new happens. Recently I went back to look at photographic prints I had made back in the early 1980s, treasured images of the house I grew up in just before it went on the market for sale. All of the photos were intact but their color was fading. Digital tools will help me restore these in one form or another, but what a tangible reminder that what had seemed a permanent souvenir is all too easily lost.

Now I’m wondering about my old movie collection. I have my share of DVDs in various formats including Blu Ray, but over a long period I used a VCR to tape countless film noir and other movies from the 1930s and 1940s. I’ve got the original collection but it will be rendered unusable when my VCR dies because the Japanese company Funai Electric has now stopped making the machines, and Funai Electric was the last VCR manufacturer still standing.

The funny thing about VCR tapes is that they have a fairly long lifetime, but only if you don’t watch them. I have a few films, like 1948’s “Out of the Past” or 1944’s “To Have and Have Not” that have been played so many times the picture has sharply deteriorated. DVDs are certainly better, but that means replacing every film, a rather expensive proposition I’d like to avoid.

And remember Zip drives? I had one for a good long stretch and now have about 25 Zip disks that are useless without a Zip drive to read them. Hmmm…

Maybe Vinton Cerf has it right. The founder of the TCP/IP protocols that make the Internet possible, Cerf calls for digital vellum, the latter being a parchment made from calfskin used to create documents since Roman times. Cerf’s idea is to create snapshots not only of our content but also standardized descriptions of the devices needed to use it. We can’t ignore the reading device, because finding an old VCR tape without the player is pointless. How many old computer data tapes are all but inaccessible because of a lack of the necessary reader?

The problem generalizes throughout our digital environment, and is intensely studied by librarians, archivists and people with a passion for retaining what we have. The same impulse in non-digital form drives the Svalbard Global Seed Vault, a Norwegian installation that can hold up to 4.5 million seed samples, meaning almost every variety of food crops on the planet can find a home there. The plan is to keep the seeds safe for centuries, perhaps for millennia.

Consider the Global Seed Vault a hedge against global catastrophe, and a reminder of what can be done. We need better data migration strategies. NASA erased its footage of Neil Armstrong and Buzz Aldrin walking the lunar surface during Apollo 11, and was forced to restore the imagery as best it could from archival television backups, a severe compromise but the only restoration possible. When data are priceless, we need to get out of the habit of making such blunders.

Our hedges against digital catastrophe are being developed in places like the Digital Preservation Network (, one of a group of cooperating entities creating a network of storage systems to ensure against failure in any one node. The archival impulse is strong, but the growth of new data is overwhelming. How we preserve that data and retain a complete picture of how to access it in whatever format is one of the most pressing issues of our time.

Paul A. Gilster is the author of several books on technology. Reach him at