How I Was Born 300 Years Ahead Of My Time

This is an editorial I wrote for the April, 1977 issue of the world’s first economically successful personal computer magazine publication, BYTE magazine. The idea of the magazine was mine developed over the years 1973-1975. In an exciting entrepreneurial era, with a partner who knew the advertising supported newsstand/subscription print magazine business as well as I knew computer hardware, software, and writing we successfully built BYTE magazine from scratch starting in May of 1975. I was editor of BYTE from its inception in May 1975(first issue September) until the end of 1980 after we had sold our BYTE Publications, Inc. publishing company (i.e. the Peterborough NH based magazine and its publication enterprise) to a major New York City publisher.

Fast forward to 2006. After BYTE, I founded my (initially 100% personally owned eponymous company) Helmers Publishing, Inc. After starting and publishing three further print then print with WWW publications, it was time for me to move on. Circa March 2006, I signed the agreement to transfer my remaining ownership of that company to my co-owner in return for a modest payment and transfer of the www.helmers.com internet domain to my personal ownership.

One of my goals for my WWW site is to put some essays of historical interest about my participation in the transformation from an era when computers were kept in logical “fishbowls” (air conditioned “mainframe” computer rooms that still exist for internet WWW servers all over the planet) to the 21^st century model of desk top / hand held / lap top personal computers connected over the internet. The first step is to recover computer versions of selected writings I published in the late 1970’s such as this editorial and related correspondence with Arthur Clarke over the years.

On November 4, 2006 I scanned the individuals pages of my published editorial into image files using my current Epson 3590 scanner. In order to have flat sheets to scan, I took a spare original copy of that April 1977 issue of BYTE, and – horror of horrors to physical artifact librarians -- I cut off the selected copy’s binding using my wood shop’s table saw, its fence and a couple of sacrificial scrap sheets of ¾” plywood to hold the magazine in alignment between fence and blade.

After I scanned the individual pages that contained this text into a bunch of 1200 DPI “.jpeg” intermediate files, then on December 15, 2006 I used the “AbbyWord” OCR software that came with the Epson Scanner to convert these pages into .txt files. Then I inserted each of the scanned segments as text into this MS Word .doc file. After aggregating the OCR output as one file, I word for word proofed the new computer file against the original printed source.

On March 21, 2007, I did a final conversion of the file into the form you see here by deleting the hard returns that I had inserted at the end of ech original published line so that the resulting Open Office Word .odt file would make a nice neat HTML file for posting on my WWW site...

It started late last summer, the first of the rumors and innuendoes. Someone walked up to me and said words to the effect of "you know, Carl, Arthur C Clarke writes well about you. . .." How would you feel if someone walked up to you and made such a profound statement? To an old science fiction addict turned science fictionish reality purveyor, it is a rather astonishing statement. Arthur C Clarke is sitting up there in my private pantheon of the gods of great literature along with Robert Heinlein, Poul Anderson, Isaac Asimov, and Larry Niven (as well as editor John Campbell, now deceased, who cultivated the talents of many science fiction writers over the years). I've never spoken to the man, communicated with him, let alone imagined that I would. He's the person who suggested the idea of the synchronous satellite before today's technology ever existed, and is author of numerous widely read science fiction stories.So when someone suggested to me that one of those remote names on book jackets had somehow taken notice of yours truly, my immediate reaction was to file it as amusing and forget it, as if someone had walked up to me and said "I understand Herman Melville had a whale of a good time with you when you guys went to school together."

However, this vicious rumor kept dogging me. More than one person mentioned it in passing. Eventually I found out, another example of the truism that if there is a persistently rumored or mentioned fact, there is often a source to that story, whether it be a persistent gossip columnist or a more pleasing manifestation of reality. It seems that according to Arthur C Clarke, I really lived in the 23rd century of our current calender[sp]: That noted historian of the future wrote a historical novel about my adventures as the mildly tragic hero of his latest book on life in the 23rd century, Imperial Earth. The book is available at any local book store in paperback form, on the science fiction shelves. It's Clarke's contribution to the American Bicentennial. Like most historical novels (according to my mental stereotype), it is a mildly boring book which takes a lot of pages and features a slowly developing plot, giving ample opportunity to tour the wonders of the age in question. I find on reading the book that one of the three major characters in the story is named (at various places — the editors were inconsistent) Karl Helmer or Karl Helmers.

What a shock. While I would have liked to imagine myself the principal by namesake of an Ayn Rand novel, I guess I'll just have to put up with Arthur C Clarke as a distant second best. (I really have no choice in the matter, and I'm certainly not about to be so superficial as to change my name to John Galt.) However, it is an interesting coincidence since there are numerous parallels with the character in the story (as well as vast differences). Here I am, fancying myself an imaginative technologist in an age of technology, reveling in the fun I'm having, and all of a sudden I find that the hero of one of my own hero's novels is an imaginative technologist in another era, with a very close approximation to my own name. Aside from the unfortunate end of the hero's life in the story, the net effect of reading the novel was an ego boost I could not in all modesty ignore.

Science fiction occasionally catches up with reality. (See Henry Melton's article "Why Aren't There Any Altairs on Arcturus IV?" in this issue.) Arthur C Clarke's Imperial Earth presents a concept of the personal computer which is in effect an evolutionary outgrowth of what we see now, but takes the pocket calculator (programmable) as his point of departure. The name of the beast is the "Minisec," a sort of advanced pocket calculator, which is introduced in chapter 18 when the principal character of the story (Duncan Makenzie) reaches Earth on his pilgrimage to the 2276 celebration of the Quincentennial. The specifications of the Minisec? Read the following excerpt from pages 106-107 of the Ballantine Books paperback edition:

The 'Sec was the standard size of all such units, determined by what could fit comfortably in the normal human hand. At a quick glance, it did not differ greatly from one of the small electronic calculators that had started coming into general use in the late twentieth century. It was, however, infinitely more versatile, and Duncan could not imagine how life would be possible without it.

Because of the finite size of clumsy human fingers, it had no more controls than its ancestors of three centuries earlier. There were fifty neat little studs; each, however, had a virtually unlimited number of functions, according to the mode of operation—for the character visible on each stud changed according to the mode. Thus on ALPHANUMERIC, twenty-six of the studs bore the letters of the alphabet, while ten showed the digits zero to nine. On MATH, the letters disappeared from the alphabetical studs and were replaced by X +, / —, = and all the standard mathematical functions.

Another mode was DICTIONARY. The 'Sec stored over a hundred thousand words, whose three-line definitions could be displayed on the bright little screen, steadily rolling over page by page if desired. CLOCK and CALENDAR also used the screen for display, but for dealing with vast amounts of information it was desirable to link the 'Sec to the much larger screen of a standard Comsole. This could be done through the unit's optical interface—a tiny Transmit-Receive bull's-eye operating in the near ultraviolet. As long as this lens was in visual range of the corresponding sensor on a Comsole, the two units could happily exchange information at the rate of megabits per second. Thus when the 'Sec's own internal memory was saturated, its contents could be dumped into a larger store for permanent keeping; or, conversely, it could be loaded up through the optical link with any special data required for a particular job.

The Minisec system's characteristics as envisaged by this passage are not really so much science fiction. Aside from the size constraint and memory capacity, the functions of the Minisec as described can be accomplished by a well designed contemporary personal computer system. The char-acteristic of fitting into the size of a hand held object is not even necessarily a valid engineering goal for such a powerful device.

Have you ever considered the problem of typing text passages into a keyboard the size of a calculator? Aside from the frequency and positioning idiosyncrazies of the standard Roman alphabet typewriters we all use (which could be solved by training with the Dvorak system), the design of the text keyboards is well adapted to the problem of quickly typing text. I know that personally from the immense amounts of keystroking I do, and anyone who works with computers at a software level will have similar experiences. A major portion of the information processing operations performed with the personal computer are now and will always be accomplished using such text keyboards, at least for the foreseeable future, barring invention of really efficient voice parsing techniques. Thus the physical size constraint is now and will probably remain at the level of the size of a text keyboard's physical layout, hardly a hand held size, but certainly a lap-perched device.

Redefining of key interpretations is already accomplished automatically when a command language is used with symbolic commands typed as text strings. However, having a changeable command keyboard such as Clarke describes is a neat idea. I ran into a specific case of such a keyboard when I worked for a short time on an experimental digital avionics system design which employed projection displays to define different legends for key positions depending upon mode of operation during an aircraft's mission. It is a nifty concept, but a bit expensive at present. The model of the Minisec keyboard is a very sensible extrapolation of liquid crystal style display technologies coupled with switchless capacitive touch contact zones.

Forgetting the physical characteristics of the implementation, the logical characteristics of multiple modes of operation are quite realizable with present day personal computers, although not always in a prepackaged form as described by Clarke. For example, the whole purpose of the ALPHANUMERIC mode is to allow text entry and command language style operations, if it is meaningful to have such a mode at all. The ASCII keyboard and the text editors applicable to personal computers of today provide this mode of operation. The MATH mode is simply extending the concept of a BASIC or APL or language X interpreter to allow its use with your personal computer to calculate programmed combinations of elementary and transcendental functions. DICTIONARY mode is a little bit more difficult, primarily because as of today we do not yet have the technology to inexpensively store (with random access) such a large vocabulary with associated definition materials. But, given a mass storage drive with 2 million or so bytes on line (for example, a 3M cartridge) and enough time to do the keystroking, a personal information system (a la dictionary) with a significant size can be built using equipment presently available on the market. The CLOCK and CALENDAR functions of the Minisec are so trivial to implement that it's amazing it isn't standard on all personal computers which are assumed to be plugged in 24 hours a day: It only takes dedication of an interrupt line driven by timing signals generated by frequency division from the crystal controlled clock of the processor. The interrupt routine updates the real time as stored in an appropriate programmable memory location or locations, scaling and converting to 24 hour time if desired.

As for connection to external larger information systems, the present day personal computer does not exactly use a fast light interface, but it certainly can use a modem and telephone link at 300 bps or so. Modems are available in kit or surplus form at quite reasonable prices today, and are likely to be a drop in the price bucket compared to the costs of the timesharing services themselves unless you just talk to your neighbor's computer.

What this all boils down to is the conclusion that the design and function of Arthur C Clarke's conception of the ultimate personal computer is a lot closer to reality than he might expect. The actual dates he gives for the progress of technology into the Minisec stage I do not recall, but I certainly suspect that he put it further into the future than the present day and the next few years. But, then, one purpose of science fiction is to explore technological scenarios given an author's defined set of outrageous premises. No one should believe that there is any prophecy about when the scenarios are played if ever. Rather, the ideas and speculations serve as inputs to the imaginations of the innovators of practical products as the technologies progress. Viewed in this light, Arthur C Clarke's concept of the Minisec personal computer and its uses throughout the book Imperial Earth is a most inspiring input which designers of personal computers would do well to explore."

Occasional, functional specifications of neat ideas can be fun. Here is one which was dreamed up by yours truly and Judy Havey while preparing the April issue. We both love to listen to classical music FM radio stations while working, principally WGBH and WCRB out of Boston MA. What we were talking about was the fact that while we certainly appreciate the fact that advertising supports the commercial radio station WCRB, we sometimes get just a bit tired of hearing the same commercial or announcement over and over again. So, I said to myself, wouldn't it be a neat hack to develop an audio input processor backed up by an AI algorithm which has the following characteristics:

• Continuous digitization of the signal and characterization of the past “n" milliseconds by a set of numerical constants reflecting the signal being digitized.

• A manual switch input used for "catching" various unwanted signals and storing the empirical parameters for the previous time period (goal: perhaps 16 to 128 bytes characterizing a segment of the signal with a high probability of uniqueness).

• A monitoring algorithm which seeks a match between stored signal parameter sets and the incoming pattern, perhaps with a "criticality" parameter which adjusts a fudge factor tolerance on the match. When a match is found, the "squelch" signal is given to the output level control peripheral.

• A level control peripheral which is either full "on" or in a squelched state, with a timing oneshot used to stretch the squelch period over an adjustable time of a half to two minutes or so.

In this simplest conception, no attempt is made at fancy word recognition, or recognition of voice characteristics of announcers.Such embellishments would of course be most desirable, but who knows what is possible until it is tried? Who'll be the first to build this logical squelch device and write it up as an article for perusal of our other readers? . .. . CH

The following is a letter to the editor from Arthur Clarke that I published subsequently in the August, 1977 issue of BYTE magazine. ...CH

Yes, it's a small universe . .. and I'm embarrassed to say that I'd never even noticed the name of BYTE's editor. . .

Although the January issue is the latest to arrive here by balsa-wood raft, my ever-loving editor at Ballantine Books, Judy-Lynn del Rey, has airmailed me your editorial on Imperial Earth, which I read with great interest.

I was well aware of the total impossibility if imagining 23rd century technology, and am glad that my "Minisec" isn't already listed in next season's models. Incidentally, the new Casio CQ-1 does contain a clock/calendar of the type I suggested!

However, the COMSOLE is already upon us! Earlier this month I watched the British Post Office engineers demonstrate their VIEWDATA system at the Venice EUROCON 77 (see enclosed leaflet).

When fully developed, it will do essentially everything I described in Chapter 24. Concerning alphanumeric keyboards for a hand-sized device, I agree that you wouldn't use them for extended keystroking. But you'll need them to make brief entries for such purposes as directory and dictionary information (vide the problem of the wristwatch telephone directory, which I e centennial —you'll find it in the upcoming The View From Serendip).

When I was in Boston last year I was fascinated by the video displays in Marvin Minsky's lab, and am determined to take one back to Sri Lanka with me on my next visit in October. . . if I am able to make a choice from your embarrassing selection of riches. (ECD's Micro-mind seems the one that best fills the bill, as all I really want to do is to make pretty pictures and impress the neighbors.)