I saw this post about the Digital Equipment Corporation’s PDP-11 mini-computer on Ars Technica and it brought back a flood of unexpected memories. For some people, a song from their youth brings forth a flood of memories, for me, my madeleine de Proust is a computer. And, strangely, not even a computer I ever owned, but one that I desperately wanted. An unrequited love affair with a machine… Well, maybe that is going a little too far… But a version of this computer set me down a career path, another version of the same computer caused me to take a hard detour that indirectly led to my current career.
Some background.
It is the 1970s. My dad is an engineer at Texas Instruments (TI). We always have the latest calculators as they are released. Or at least the ones released by TI, especially the “SR” series meant to be electronic replacements for the then ubiquitous slide-rule. Take a moment to consider what a game changer the portable scientific calculator represented. Before 1972, when the TI SR-10 and the competing HP-35 (with its unusual Reverse Polish Notation, or RPN, entry system) were introduced, scientists and engineers, and students of the same, all carried slide-rules to perform quick computations when needed. The other options included booking time on a shared mainframe computer system (remember, this was before the days of personal computers), or use an unwieldy desktop computing device (or, do the computation by hand). Starting in high school, students who were planning to advance in technical subjects were expected to master slide rule operation with teachers sometimes using very large demonstrator units. My father even worked as a slide rule demonstrator in college for a time. Engineers could often be spotted with their pocket protectors and slide-rule holsters on their belts.
The new easily portable calculators changed all of that. Suddenly, slide-rules were out and calculators were in. TI, which had been heavily focused on semiconductor manufacturing, RADAR systems, and government contracts, suddenly had a booming consumer division, and was launching an array of new products including watches, thermostats, and of course, the iconic Speak-n-Spell. During this time, a computer was either a time sharing mainframe, or one of the newly created minicomputers. TI had one, the TI-990 but personal computers were on the horizon. The MITS Altair kit was released in 1974, Bill Gates read about it, developed an implementation of the BASIC programing language, and Microsoft was born. IBM released their APL based portable computer, the IBM 5100 in 1975. So the nascent computer revolution was in the air in the mid-1970s, and also in the air in our family home..
My story continues
During this time my dad had several hobbies. His primary interest was photography. He ran a small photography business on the side, and we built a professional darkroom in the garage. He also enjoyed building electronic kits, particularly Heathkit kits during this time. He would sometimes let me help him with his projects, and I learned how to use a soldering iron, how to read resistor color-codes and follow assembly instructions for the various kits he built.
So, it is the the mid-1970s my dad is building electronic kits, bringing calculators home, talking about how computers will change our lives. We are getting catalogs from companies like Paia (advertising theremins and modular synthesizer kits) and Heathkit, and others. While my grandmother combed through the Sears catalog, I read all of these electronic kit catalogues cover-to-cover looking for interesting projects to suggest to my dad. One year when I was 10 or 11 (so it must have been 1977 or so) Heathkit ran a multipage spread in the catalog announcing their latest kit. A computer! And not just a “microcomputer” like the Altair, but a real minicomputer, a kit version of DECs powerful PDP-11, but in a smaller case, it was the debut of the H-11, a16 bit computer that you could build and program in your home! and unlike the Altair, the H-11 actually had a ready-made eco system of software and hardware that it inherited from its DEC siblings.
I remember reading the description in the catalog, not only of the H-11, but of all of the devices that could be connected. The terminal, the printer, the really slow (was it even 300 bps?) modem with acoustic coupler, the paper tape reader. I think there may have even been a disk drive available! I was only 10 or 11 years old, but I knew that I really wanted a computer! I starting reading books and magazines about computing. I already enjoyed science fiction, so I was primed to think about all of the ways that I could use a computer. I even took a typing class in elementary school (that used old fashioned manual typewriters, no less) so that I could be able to use a terminal. I went to bed with the Heathkit catalog in my hand, dreaming of someday owning my own H-11!
Unsurprisingly, I was not able to convince my father to buy such an expensive device for our family. For one thing, it cost over $1200. That is a lot of money now, but much more back then. This was before personal computers were much more than a techno-fantasy concept (the weird and wonderful IBM 5100, not withstanding). However there was talk of the future when every home would have a computer, but it was not really clear yet how such a device would be put to use in the home. Word processors would come a few years later, and Dan Bricklin had not yet invented the spreadsheet application. The applications envisioned for home computers included home recipe management (but my mom did not cook, so that was not an issue), home finance, household programming (i.e treating the computer as a calculator), education (we already owned a Speak-n-Spell and a Speak-n-Math) and entertainment (but we already had electronic games like Simon and Merlin). Also TI was working on a “Home Computer,” that my father assured me would be far superior (this was the TI 99/4A, which did end up being my first computer and which was superior in someways to the PDP-11 for home use, and certainly could have been superior to the IBM PC, but ended up being hampered by TI’s own lack of vision, including a belief that people wanted to use their home TV as a monitor and that average home users would not be interested in buying a disk drive, but that is a story for another day)
So, we did not get an H-11. The next year, I went to a new school, and while there was no PDP-11, there was an Apple II! I learned Apple Basic. A few years later, TI produced an over supply of TI-99/4As so my father dutifully bought one with all of the accessories at the employee store at a deep discount. I played around with the Extended BASIC module, got really fascinated with the FORTH systems, took an assembly language for kids class (TMS 9900 assembly language of all things!), made some awful noise with the 4-note polyphonic synthesizer with programable ADSR, learned about sprites and other geeky computing things. I took computer math, then AP computer science (the first year it was offered) and finally a second year of computer science in high school where we used the same textbook (SICP) that MIT used to learn the Scheme dialect of LISP. I read Gödel, Escher, Bach: An Eternal Golden Braid. I joined a Computer Explorers club hosted by Convex computers. I got to play with a few weird computer systems, including the TI Explorer, based on the MIT LISP machine. Also, during this time I first learned to use TeX.
My first job (at 14) was working at the TI store in the Prestonwood Mall demonstrating and selling TI 99/4As (and watches and calculators). In the summer after graduation, I worked for a typesetting company, writing SNOBOL4 programs to massage text and data from various sources (including 9 track tape drives) into a format usable by our phototypesetting system. I was on my way to a fruitful career in technology. I won a Presidential Scholarship to New Mexico Institute of Mining and Technology to study computer science.
A jump to the left…
I started university with a firm background in programing, some exposure to the more theoretical side of computer science and some knowledge of the contemporary state of the art in artificial intelligence research. I placed out of all of the introductory programming classes, and started my first semester with a fairly heavy load of CS classes, including systems programing, programing languages, and practicums in Lisp and Prolog. I was excited to take the system programming class, as it utilized the department’s VAX 11/780, essentially an upgraded version of the PDP-11, the bigger version of the H-11 that I wanted so badly as a child. We got off to a rocky start. Rather than teaching LSI-11 assembly language directly, we were to write an assembler and linker in C, that could then create an executable file based on an assembly language program provided as input. From a CS perspective, I suppose this makes sense in a way. Basically it is writing a compiler for a well defined language, learning how link loaders work, and how to make calls to the operating system (BSD Unix, in this case), etc. But, I found that I wanted to learn the code first, not how to write other code to transform the and execute the assembly source. I suppose that I just did not understand at the time what “systems programming” was about. I ended up spending about 20 hours per week on this single 1 credit lab class. I did not have a way to remotely access the VAX, so I had to go to the computer lab to complete my assignments. There were very many stressful late nights spent in front of terminal debugging my C code cursing the unseen machine down the hallway. At the end, I barely passed, and went through a period of loathing for the former object of my desire.
…and a step to the right.
I eventually transferred to a liberal arts school, losing interest in computing for a while. I did take a work-study job in the computer lab at my new school, but that was mainly helping other students use Word Perfect to write their term papers, maintaining the printers, troubleshooting PCs, and doing general desktop support. I also took an automata theory class, a discrete math class, and a linear algebra class during this time. In fact, I enjoyed my automata theory class so much that I very briefly considered going back to engineering school once again to complete a computer science degree. I eventually graduated, moved to Colorado, and started graduate school in Religion (pretty far afield from Computer Science). But, I needed a job, and there was a position listing for a back up tape administrator. The responsibilities involved coming in at night to load tapes into the 9-track tape drive, clean the drive occasionally, and do other small systems administration tasks as needed. That worked out well, but eventually the backup part of my job was replaced by a tape robot, and I unwittingly became a Unix Systems Administrator for an HP-UX system. That was followed by a brief interest in database administration and SQL report writing, when I started asking questions about our new Cisco router, and before I knew what had happened, I became a network engineer, working for a dynamic and growing internet service provider at exactly the right moment just as the industry took off.
Now I am about 30 years into a career in telecom. If I had not had that first encounter with the PDP-11 / H-11, I might not have been interested in computing. Had I not had the negative experience with systems programming on the 11/780, I might have stayed in computer science or programming and missed out on networking and the Internet. So, I am happy where I am and grateful that I encountered that catalog listing for the H-11 all those years ago.
