You know. Russian Nesting Dolls.

That’s what we’ve got for a file system. First you crash a computer. But save its files. Do it three times. Then get a new computer and move the files. But restore the files you’re working on. Do this through Dos 3.5, 5, Windows, Windows, Windows—not even mentioning IASC art programs and Photoshop, Eudora, Microsoft Office, OZwin, Golden Retriever, Compuserve, XYWrite, Wordperfect, .txt, .doc, .let, .tmp, .mem, .log, etc, etc, etc, from 1980 to 2009. Nest all of that. Then, starting with the 1980’s, realize your 5 1/4 disks are outmoded, so load all your backups onto the machine that can still read those, including the floppies from the luggable’s drive, then convert to 3.5 stiffies, and have a system crash. Back up again. Save everything. Get it on disks. Oops. Tape backups, IOmega backups; get all those onto disk: they’re going to change the media again. Laptops. Another version of windows. Internet downloads. Webpages, in all their versions. Frontpage. NAMO, htm, html. System crash, re-backup. Now everything is going to cd-rom. Reload everything. Back up again. Motherboard dies. New system. Bring up old data. Finish getting it onto cd-roms. Oops. Now we’re DVD. Get everything backed up off 3 laptops and the desktop which is the central repository of old files.

Then you stare at a stack of dvd’s that could populate a store, and realize—it’s a mess. It’s a real mess. Everything is backed up, but there are no labels that make any sense: a cd or dvd just can’t list everything that’s on it and you’d have to boot each to find out what’s there.

Sigh. And do it for two and three writers under the same roof, interchanging files and using, at times, the same machines, then linking by house network.

This is where we threw up our hands and got the terabyte drive. I’ve put the thing in my Amazon store: costs 128.00, and is thus far immaculate. We have shoved everything since 1980 onto that drive, and we find directories that look like this:

Directory A: directories B-N, another copy of directory A INCLUDING B-N, and loose files that belong to directories O-Q, and one to B, plus some unidentified .exe files. Inside the second director A is, of course, another Directory A, and all those files and directories. Repeat this for 20 levels.

Ever been in a hall of mirrors? And some of the files differ from the ones in the later (enveloping) directories by a handful of bytes. Others differ by hundreds.

I am going down that list. I shudder when, embedded in Directory A 10, I find a file labeled: CJC Backup files, which contain, yes, MORE Directory A’s. Or “Misc Directory”—an adventure involving stray single files and a couple of unique subdirectories.

My technique. Every directory I’m working with gets renamed 1AAname. This throws it to the top of the heap. Once resolved, its pieces are renamed Zname, which drops them to the bottom. If a directory contains extraneous files or pieces whose parent directories are z’ed and out of reach, I whisk those into a directory named 1AAMisplaced, which I later empty. You ‘steer’ this mess by renaming either 1AA or BB or zz or the like. And once you get a collection of files with similar names all together in the list, you run Duplicate File finder on that set, and it will dump the chaff, and id the ones that do need changed names to coexist in the same file. So finally you have one pure Directory A that has only Directory A files, and Directory B is beginning to shape up, etc, etc, with its own unique files. I’m taking the little matroushka dolls out of the larger dolls and am creating order in the universe. This is actually going faster than you’d imagine.

Let me give you some tips from my bad example. 1. If it is likely you will want to preserve unique versions of your files, use the date as part of the file name. A lot of these were done while we were only alloted 8 letters for a file name. Take advantage of the additional letters, and call things by unique names. Twenty-two files of varying length all named EXPLOR1.wpd are not helpful—particularly if the computer helpfully suggests it can overwrite anything of the same file name to solve your problem. You would not want that to happen. No, nay, no way. (2. do not leave loose unrelated files in directories that have other directories in them. (3. Do not endlessly back up nested directories: they only get worse and deeper with every pass. Unravel the nests before backing up, and label so you have a clue what’s actually backed up. A directory full of novels that is already 3 directories deep should not be labeled “Archive” and blindly backed up with every new drive you buy. IE, do not make your working hard drive the ultimate repository of massive files. Just back the sucker up and shelve it. My life would be so much simpler if I’d done that.

Of course, as fast as they were changing media formats I had a paranoid fear of seeing my work turn out unreadable, the way my old Atari floppies did. Though I got almost all of them (late 70’s). The cellwriter and typewriter, alas, cannot be resurrected except by scanning.

I love this Duplicate File Program.