Next post Previous post

Text files and me - Part 1

February 15, 2011 at 10:59 PM by Dr. Drang

A couple of days ago, David Sparks of Macsparky and the Mac Power Users podcast wrote a post about how he tries to do all his writing in text files. It seemed like a good idea to write about how I came to be doing the same thing. As I started writing, I realized this is going to be long, so consider this the first installment of I don’t know how many.

First, a digression. David uses Simplenote as the “backbone” of his workflow and describes it this way:

SimpleNote is a Web based text service that grew out of an iPhone notes app but became much, much more. SimpleNote does nothing more than sync text files with the kind of focus usually saved for teenage boys at the beach. It just works. Think Dropbox for text.

Now, I have nothing against Simplenote. Like many people, I bought it shortly after reading John Gruber’s description of it (it didn’t become a free app until months later). I was, in fact, an award-winning user of Simplenote. But “think Dropbox for text”, isn’t thinking at all. Dropbox is Dropbox for text. If you want to share text files between your iPhone/iPad and your computer, it’s much simpler to use an iOS editor that accesses Dropbox—which I assume you already have on your computer—directly than to use Simplenote and the various add-ons needed to sync to its bespoke cloud service. Which is why I now use Elements instead of Simplenote.1

Back in the day

My introduction to marked-up text files came in my senior year of college. I was taking a Pascal programming course2 and spent a lot of time on terminals connected to the school’s CDC Cyber 175 computer, a mysterious and powerful machine that was kept in a climate controlled room in the bowels of the Digital Computer Lab and tended to by a presbytery of sysadmins. One of the programs on the 175 was RNF, a text formatter that was—like Unix’s troff, which we’ll be coming across later—a descendant of runoff, the Ur-formatter from MIT’s CTSS project.

Like troff, RNF used dot commands interspersed with the text itself to give formatting commands. As this was over three decades ago, I don’t remember the commands themselves, but I suspect they were similar to those you can still use in troff/groff/nroff.

I wrote a couple of term papers in RNF, which probably didn’t even use much in the way of markup; the standard of the time was typewritten text, and there’s very little you can do with a typewriter other than fiddle with the margins and line spacing. I think I printed them out at a paper terminal, although it’s possible I ponied up the few cents a page needed to get the final copy printed on a Diablo letter-quality printer.

After my senior year, term papers were a thing of the past for me because graduate level engineering classes assign homework, not term papers. I did type up a few papers for my girlfriend—I was, and still am, the better typist—but they were all done on a typewriter.

What happened to the RNF files I created? They probably sat there in my student account on the Cyber 175 until inactivity led to the account being culled. It never occurred to me to save the files in some electronic format, and I wouldn’t have had a way of migrating them off the 175 even if it had occurred to me. Back then, the sole purpose of the computer was to generate ink on paper that I could hand in.

Interregnum—the word processor years

In 1985, as I approached the end of my graduate studies, I bought a Fat Mac through a discounted university purchase program. I wrote my thesis in MacWrite, with figures and equations created in MacDraw, and printed it out on an ImageWriter. For many years I kept the diskettes—400 kb, single-sided, double density—that held the thesis and its backups, but eventually they were misplaced. I didn’t feel to bad about losing them, though, because they were valueless: by that time there was no MacWrite anymore and only a collector of old computers and software could have resurrected the files. I still have an original paper copy from the ImageWriter as well as a printed copy from University Microfilms.

I moved on to an academic position at a small Big Ten school, where I got a Mac Plus and Microsoft Word. Several papers were written, first in Word 1.x and later in the notoriously bad Word 3. At some point, I switched to WriteNow, which didn’t have the long list of features Word had but actually worked the way it said it would. Everything I wrote in Word and WriteNow now exists only as paper, mostly in bound journals in university libraries (although it’s possible I still have a few preprints squirreled away somewhere). Again, even if I still had the diskettes with the electronic files, I couldn’t do anything with them without the help of a computer archivist.

I started working for an engineering consulting firm, where I wrote lots of reports for clients, first in WriteNow, later in Word 5, which I started using only after colleagues convinced me that the stink of Word 3 had worn off. The electronic files for these reports were left behind on my company-issued hard disk when I quit to start my own firm. Even if I had taken the files with me, I doubt they’d do me any good now—orphaned formats again.

The new company started on a shoestring, and I decided Word was too expensive. (Also, Word 6 had come out and everyone said it wouldn’t run on lower-end hardware.) So I switched to Claris Works for a couple of years. Yes, the electronic versions of those files are gone, too.

This brings us to 1996. Apple was in crisis, and rightly so. The Mac OS was in terrible shape: it didn’t do real multitasking (remember cooperative multitasking? remember choosing how much RAM a application would use?) and crashed frequently. When Apple announced that it was looking to buy a company to write its OS—an admission that it really had no idea how to fix its own product—I bailed.

The revelation

At this point, I had been generating papers and reports on computers for over 15 years and had, electronically, nothing to show for it. Everything I’d written was either missing or as good as missing because it was in a format I no longer had easy access to.

In the last months of my Mac years, several things came together to put me on the path I’d eventually take:

  1. I’d been learning HTML and CGI and making web pages. I had a copy of BBEdit—version 4, I think—and was pleasantly surprised to see how easy HTML was. The tags reminded me of RNF, which I hadn’t thought of in ages.
  2. HTML led to SGML, a more flexible markup system.
  3. I’d been playing around with MkLinux, which wasn’t a true Linux because it used a microkernel, but still gave me a sense of what running Linux would be like. My nascent interest in web site building had given me a sense of the power of Unix.
  4. Looking into Unix led me to troff, TeX, and LaTeX, markup languages that were similar to HTML, but meant to produce printed copy instead of web pages.

I was tired of the continual word processor changes and the loss of work to orphaned file formats. I knew plain text files would last forever. I hoped I could leverage what I knew about markup languages into a writing system that would allow me to work in plain text while still producing the nicely-formatted paper reports my clients needed.

In our next episode—which may not be for a while—I’ll explain my move to Linux and the workflows I built to satisfy both me and my clients.

  1. In fairness to the Simplenote guys, I’m pretty sure the reason they built their own syncing service is that Dropbox hadn’t yet opened up its API when Simplenote was designed and written. Later iOS editors like Elements, Plain Text, and iA Writer had the advantage of coming along after Dropbox syncing was available to anyone. ↩

  2. From this guy who, unlike my other professors, is not yet emeritus or dead. ↩

Next post Previous post