Sometimes all you can do is make periodic manual backups and hope you don’t lose too much

EgyptScribe4Ever since a hard disk crash in 2009 when I lost four months of work, I’ve been a fanatic about having constant up-to-the-minute backups. My old system was to make periodic backups on CDs whenever I remembered, maybe every six to twelve months. But that doesn’t cut it for me anymore, because who wants to lose 6-12 months of work? Since then I’ve had two major systems. The first was Dropbox: I stored everything–all my documents–in a Dropbox folder. But Dropbox is not a good backup system, for reasons I detailed in a previous post which I can’t right now find (basically, Dropbox is designed to keep files in sync, which means if they’re deleted from your computer, they’ll be deleted from the cloud, too! Which is exactly the opposite of what you want in a backup system!!! They advertise that they can recover these deleted files, but in practice the interface is very clunky, and it’s almost impossible to recover large numbers of deleted files, which means if you suffer some calamity that wipes out an entire folder, you’re screwed.)

I since switched to Backblaze, which is a pure backup utility. At any given time it has a snapshot of all your personal files (movies, music, documents, photos), and you can have them send you a snapshot from up to four weeks prior to your latest backup, so if you delete files, you can just revert to an earlier version of their backup.

Anyway that’s good for files I host on my computer. But what about those I don’t? Right now I have data within three clouds: Amazon’s, Apple’s, and Google’s. These files are not under my control. They’re not hosted on my computer. So how do I back them up? In particular I’ve been worried for years that my gmail will get wiped out somehow. Amazon also has a practice of suspending accounts for very specious reasons (too many returns) and I’ve been worried about not being about to access the thousands of dollars of books I’ve purchased for the Kindle. Then there’s WordPress. I have well over a thousand blog posts hosted on wordpress.com. Not to mention my tweets and my facebook statuses, which could similarly go into the wind. I keep a personal journal on Evernote as well: tens of thousands of words about my day-to-day life.

I’m not worried about this stuff for tomorrow or next year. But Facebook and Twitter and WordPress or Google will not be around forever. In ten or twenty or fifty years, they will be gone, and I want to still have a record of what I did on those sites. What would be ideal would be if all of these sites also kept a synchronized copy of their data locally, so I could backup the local versions of the files. But they don’t. For obvious reasons. They want to control my files.

Which leaves me with not a very good backup solution. Basically, my practice is to, whenever I remember, manually export an archive from each of these sites and then save it onto my disk, where it then gets backed up (I also have an external hard drive that takes Time Capsule backups). But of course this is exactly the same system that failed me in 2009! Sigh. The only solution is discipline: monthly or bimonthly backups. But I’m not sure I’ll ever get it together enough to do that.

In fact, even establishing a baseline backup has been hard for me. It’s only today that I downloaded, for the first time, my Gmail archive (11 gigs). And I also downloaded all my Kindle books, stripped the DRM, and exported them into a 3rd-party ebook management platform. If Google or Amazon had fucked me up yesterday, I would’ve lost everything. But if they fuck me up a year from now, I’ll only have lost a year’s worth of data.

Progress.

(You’ll note that I don’t plan for END OF THE WORLD backups. If in fifty years we no longer have computers that are capable of reading, say, .doc files, then I am screwed. But somehow I don’t think that’s a problem. Using emulators we’re right now able to open some pretty old files. And even if file architecture–or our definition of what constitutes a computer–were to radically change, I think there’d still be someone who developed a solution to look at old files. As long as I have the thing itself, and as long as I keep transferring it onto fresh and moder storage media, I think on that score I’ll be fine.)