Future-proofing your data
Six months away from completing his sociology dissertation in 1992, Robert Runté realised his floppies were readable only by an Osborne computer – and he was the only person he knew who still used one (released in 1981, the 11,3 kg Osborne is considered to be the first “portable” computer).
So he uploaded his file to the University of Alberta’s mainframe – what could be more secure than that, he thought. Sure enough, about a year and a half later, his ancient computer died, but because he had a backup, he didn’t mind when the people at the repair place “all laughed uproariously at the thought of trying to find parts” for the Osborne in order to fix it. He was covered.
Except that years later, when he went to retrieve the data, Runté found the university mainframe he had backed up
to had been torn down and replaced by Mac servers. (Luckily, Runté had his dissertation printed out.) Today Runté, a professor at the University of Lethbridge in Alberta, backs up to multiple devices and advises graduate students to send a hard copy of every new chapter of their all-important theses to the most reliable of off-site backups: Mom.
Runté’s experience points to the ultimate, inevitable problem with data storage: all interfaces and formats eventually die. Data storage consultant Tom Coughlin, founder of Coughlin Associates, calls it a fight against Nature, saying, “The laws of thermodynamics are against you.”
Such a battle makes for a hazy long-term outlook. Will your data be accessible in 100 or even 50 years’ time? Perhaps, but those data will probably be in different formats and will certainly be stored on different media than they are today. All modern-day technologies grow obsolete; either the hardware breaks or is replaced by something better, or new software takes over for the old, or both.
In 50 years’ time, you may have a computer that can read PDFs, but you might have those PDFs stored on a medium the computer can’t read. Or the opposite may happen, with data stored on a readable format but saved in long-gone file formats. The key to preventing either case is accepting the non-stop job of staying technologically up to date.
Because there’s no single, perfect digital archiving solution, the key to making our data last forever is good habits. We need to be vigilant, continually moving our data forward to new formats and keeping it on multiple devices – before whatever we have becomes obsolete or simply fails. The best protection we have against data loss is redundancy, and lots of it. William LeFurgy, manager of the National Digital Information Infrastructure and Preservation Programme at the US Library of Congress, says the most important thing everyone should do is make copies of data and store those copies in different locations.
The first of your backups should be on an external hard drive. Look for a desktop drive (desktop drives tend to be sturdier than portable ones) and make sure it has the most common input/output interface, USB. Warranties don’t assure long life spans, but they’re a good indicator of what to expect, so find a drive that’s guaranteed for at least four or five years. Verbatim’s 2 TB USB drive has a seven-year warranty, and LaCie’s d2 Quadra Enterprise Class 1 TB drive is guaranteed for five years. Shop around.
For long-term archival storage, the media most likely to last a century are gold-coated Blu-ray discs, thanks to a special protective layer, and film (negatives and video film, as well as microfilm). The promise of a century-long life span is like a good warranty, giving at least some reassurance that the product won’t bite the dust a week after you start using it. A hundred years from now, though, you’ll also need the hardware to read these media.
Unfortunately, unlike with paper archiving, you can’t just put a disc or drive into a vault and hope to read it in a hundred years. How you store the media is important; magnetic discs need to be stored away from light and pollutants such as dust and at about 20 degrees Celsius with 40 per cent relative humidity. And long-term storage on any one device or medium can leave you stranded with an unsupported format.
Never get rid of hardware (computers, drives, readers, etc) until you’ve moved the data on to the next storage media and confirmed that it can be read. Often, this means copying not only files, but also the programs, plug-ins and drivers needed to read them. Have a bunch of already obsolete media lying around – VHS tapes, floppy discs, even CDs? Rescue those files as soon as you can.
Choose your file formats carefully. Open formats trump proprietary ones when it comes to longevity. Two decades from now, an Adobe Photoshop PSD file is far less likely to be compatible with the software of the future than an open format such as TIFF or JPEG. Likewise, plain-text files (.txt) are better than Microsoft Word’s proprietary formats (.doc and .docx). Almost any type of text or image file can be turned into a PDF/A, a special type of PDF designed for archiving.
Perhaps the most critical step many of us forget is verifying that our data is recoverable. Animation studio Pixar almost lost a year’s worth of work on Toy Story 2 when someone mistakenly entered a Linux command (rm*) that wiped away the files in seconds. The company checked its backups only to discover that they had failed. Luckily, Pixar’s technical director had been working from home, so the company was able to salvage the data from her personal computer. The moral: at least once a year, make sure your backups actually work.
Finally, after following all these steps to combat fragile media, tech obsolescence and failed backups, there’s one last hurdle: making sure your data can outlive you (if you want them to, that is). For that, rely on the same methods of digital redundancy and add in the analogue – people and paper.
Let family members know where your data are stored, how to access them and what they should be looking for. And for documents you truly can’t stand to lose, store copies digitally, but hold on to some additional copies in traditional form: on paper.
|Life Span||2 to 5 years||3 to 7 years||10 years||10 years||10 to 15 years|
|Max Capacity||650 MB||3 TB||256 GB||4 GB to 32 GB||25 GB|
|Pros||Portable and cheap, and can be read by almost any modern computer. Those with gold or silver undersides last longest.||Large capacities, fast read/write speeds, can be used with programs that automatically sync files.||Low risk of mechanical failure because no moving parts, quick startup, virtually no file fragmentation.||Portable, no moving parts, can be read and written by many smartphones and cameras.||Unlike DVDs, required to meet a base level of scratch resistance, relatively cheap, can hold up to 25 GB per layer.|
|Cons||Fragile, degrades easily in extreme temperatures, can be ruined just by markers (it’s best to write on the clear centre ring).||The constant spinning of a magnetic platter takes a toll on its health, and mechanical failure is inevitable.||Expensive, limited number of times each bit can be erased (and therefore can be written).||Easy to misplace, limited number of read/write cycles.||Fragile, degrades in heat and light unless it’s a gold-coated archival disc.|