To market their new products to people who had not already spent years pining for a computer of their own, the creators of the second wave of microcomputers had to face head on the question of what the microcomputer was actually good for. What was its value, if not as a hobby plaything for self-motivated computer nerds? To answer that question, they sketched inventive fantasies about how the computer might somehow be an aid to everyday domestic life. They also harnessed the computer’s symbolic power. By 1977, nuclear power and rocketry had begun to lose the sheen of their glory days in the 1950s and 1960s—the computer had taken over as the icon of progress, the drive wheel of the still-unfolding next stage of modernity.

Some early adopters no doubt purchased computers on the basis of this promise, convinced that having a computer in the home would somehow better prepare their children for the future. Others found joy in programming a computer without having to fiddle with hardware.

But most of the non-hobbyist audience that computer vendors hoped to reach wanted to purchase software for their computer, not write it. The computer makers tried to meet this need by creating their own software libraries, but these were always limited in both extent and quality. Luckily for them, the availability of cheap, easy-to-use computers primed the pump for a large number of third-party software producers. Their most popular products fell into two very distinct classes: business software for creating digital work artifacts, and game software to entertain. To meet the market demand for each, computer makers developed two new product categories: high-end, capable business computers, and simple, inexpensive computers meant mainly for games.

Selling the Home Computer

As the new breed of computers rolled out to the market in 1977, journalists struggled somewhat with what to call them. They were evidently no longer “hobby computers,” but should they be identified by their physical components (“microcomputer”), their ownership model (“personal computer”), or their place of habitation (the “home computer”)? In practice, all three terms were mixed almost interchangeably for the next few years, although “home computer” gradually became associated with the less expensive, less capable game machines of the late 1970s and early 1980s (*BYTE *magazine editor Carl Helmers preferred to emphasize ease-of-use with “appliance computer”, but this term never got traction elsewhere).[1]

What people would do with these computers, whatever you might call them, was another matter. Journalists and marketers alike imagined appropriately domestic tasks for the home information machine: diet planning, monthly budgeting, check balancing, recipe management. Some people, somewhere, certainly used a computer to do such things, but it is safe to say that few people were clamoring to spend $600 on a TRS-80 to take over tasks done more easily with a $15 pocket calculator and a piece of paper. It was difficult to admit that, just like the pure hobby computers that preceded them, there was nothing particularly useful for most people to do with an Apple II, Commodore PET, or TRS-80 when they first released.[2]

Even the makers of these computers were conflicted about what market they were addressing. Steve Jobs wanted the Apple II to be a friendly home appliance, like a color television or hi-fi stereo. It was only at Wozniak’s insistence that it included a generous allotment of expansion slots, like the Altair and its minicomputer predecessors. The first Apple II ads inadvertently reflected this uncertainty about what exactly users wanted. It promised the ability to play the PONG arcade game, but also to “invent your own games,” to “teach your children arithmetic, or spelling,” to “manage household finances, chart the stock market or index recipes, record collections, even control your home environment.” The first TRS-80 ad was similarly scattershot, promising to be everything to everyone: “[p]rogram it to handle your personal finances, small business accounting, teaching functions, kitchen computations, innumerable games…”[3]

The first advertisement for the Apple II. Designed by a professional marketing agency (Regis McKenna) and produced in full color, it was a cut above anything seen before in the personal computer market.

Because very little commercial software existed, both ads assumed a buyer able and willing to program all of these functions themselves. Though Apple’s first ad depicted a couple with an Apple II in their suburban kitchen, the second-generation computer was not truly a home appliance ready for use by anyone. The distinctive appeal of these computers compared to what came before lay in their low price (a Commodore PET cost about one-third as much as a comparably-equipped Processor Technology Sol-20), the absence of any expectation of prior electronics knowledge, and a BASIC interpreter provided in ROM (read-only memory) so that it could be used immediately after turning on the computer, without even having to load it from tape. This last design decision gave these computers instant appeal to the many thousands who had picked up a smattering of BASIC on a time-sharing system in high school or college, but felt intimated by the work required to set up a hobby microcomputer.[4]

Commodore did little if any direct advertising in the U.S. in 1977-1978 (I have found none in the contemporary magazines available online), relying on word-of-mouth and retailer advertisements. And in a sense, they were right to let the PET sell itself, because the rapidly growing cultural force of the computer as an abstract idea also helped sell the second generation personal computers. In every newspaper and magazine, middle-class readers of the 1970s could find pundits pontificating on the dawning “information age,” “electronic age,” or even “computer age”. And they could see the refracted light of the ascending sun for themselves: by the late 1970s, computers, once hidden away deep inside universities, corporations, and government offices, had sprouted out into the everyday world, finding their way into everything from General Motors automobiles to McDonald’s cash registers.[5]

Middle-class households had also by this time become accustomed to expect the annual arrival of the latest electronic gizmo as a perfect birthday or Christmas gift: pocket calculators, digital watches (watchmaker Hamilton called theirs a “time computer”), citizen-band radios, and home video games systems. Newman Computer Exchange/CompuMart, an Ann Arbor, Michigan computer retailer, played directly on this expectation in their advertising for the Commodore PET: “Here we go again… Two years ago it was the pocket calculator. Last year it was the digital watch. Now, incredibly, it’s the computer itself.” Each of these gadgets, beyond their actual functions, also acted as an admission ticket to the information age, and nothing could outdo the personal computer in that regard.[6]

But programmers and gadget-lovers would carry the personal computer only so far. David Bunnell, editor of Personal Computing, knew that the key ingredient would be pre-written application software:

It is true that the average home computer user isn’t going to have the patience to learn BASIC, but several hundred thousand Americans have already learned or are learning the language in schools. The hope is this: give a PET type computer to every person who knows how to program BASIC and they will blaze a software trail right into the American home. They will write millions of programs that tomorrow’s home computer user can simply plug into his computer and modify to his own needs without knowing a thing about programming.[7]

Most people would only buy a computer along with software to let them immediately do something with it. In the late 1970s, this mostly meant word processors, VisiCalc, and, of course, games, games, and more games.

Word Processors

All of the second-generation computer makers knew that not all buyers would want to write their own programs, and so they offered a small library of software on cassette tape (or later floppy disk). These included simple games, programming utilities, and basic business software (such as for generating accounting reports, or maintaining a mailing list). But all of the innovations and the vast majority of sales in personal computer software came from third party authors. This was a distinctive feature of the personal computer throughout its history. In this fast-moving, highly-competitive market, no computer maker ever successfully vertically integrated downstream into application software (and only Commodore ever had upstream vertical integration into computer chips through their ownership of MOS Technology, but even they switched to a third-party supplier for the Amiga series in 1985).[8]

The new breed of software makers that arose in the late 1970s would charge for their programs without qualms. The free-software hobby culture that had booed the greed of Micro-Soft for charging for BASIC was drowned out by a much larger group of latecomers with no prior expectation that software belonged to the commons.

Given the lack of established players and minimal user expectations, the barriers to entry to software production were very low. One- or two-man microcomputer software shops sprang up by the hundreds (possibly thousands) in this period. Most disappeared again just as quickly, but a few made a lasting mark. Software publishers also appeared, who would do the heavy lifting of marketing, manufacturing, and distribution for these many scattered creators, in exchange for a cut of the sales. But this never stabilized into a consistent publisher/author relationship the way the book or record industry did; larger software publishers often brought their own development in house in order to control the intellectual property that their profits depended on, and build a coherent and strategic product line.[9]

The most significant piece of application software to appear before 1977 was Michael Shrayer’s Electric Pencil. Word processors existed as a product category already. They had evolved from mechanical typesetting systems into dedicated, self-contained computer systems from companies like Wang Laboratories, that offered fully digital composition systems with a built-in screen to businesses at prices upwards of $10,000. But Shrayer (understandably enough) had never heard of these word processors, and was not trying to emulate them. A technical obstacle, not a market opportunity, initially motivated him to create the first microcomputer word processor: the difficulty of producing the complex formatting required to document a computer program from the linear feed of a typewriter.

A semi-retired film industry veteran living in Los Angeles, Shrayer also belonged to the electronics-loving set who got hooked on computers, and became an early member of the Southern California Computer Society. Unhappy with the code editor provided in Processor Technology’s Software Package 1 (SP-1), he revised it and released Extended Software Package 1 (ESP-1). He then backed his way into creating *Electric Pencil *to free himself from the tedium of typing up documentation for all the different variants of ESP-1 he created for different hobby computer models and configurations. This itself became a year-long labor, resulting in the mail-order release of the first version of *Electric Pencil *in December 1976.[10]

But the tedium didn’t end, because despite the fact that he only supported computers with Intel 8080 processors, expanding the market for *Electric Pencil *required Shrayer to write dozens of different versions for different combinations of peripherals: Tarbell, CUTS, or North Star disk or cassette storage systems, standard or Diablo printers (the latter supported proportional spacing), and SOL, VTI, VDM or VIO video terminals. Each required custom code, because no standardized protocols existed for interfacing with these devices. Depending on the sophistication of the hardware involved, a given version of the program went for as little as $100 or as much as $300.[11]

We should not imagine that *Electric Pencil *was the functional equivalent of a twenty-first-century word processor. The ability to format text in the initial versions was very limited, and when possible, it required special command strings with the text itself: to underline text, for example, required putting underline characters on an otherwise empty line just below the text to be underlined (or above, depending on your type of printer!). Inserting, selecting, and moving text all required memorizing special keyboard combinations. Nonetheless, many tasks that were tedious or impossible on a typewriter: correcting typos, reordering words, rewriting or inserting a passage into the middle of a body of text, or finding and replacing a word throughout the entire document—became comparatively effortless on Electric Pencil.[12]

Electric Pencil running on a Processor Technology Sol-20. Note that it has no dedicated UI elements whatsoever, just text on a screen.

Shrayer disliked the 6502 microprocessor, so although he produced a version for the 8080-compatible TRS-80 (despite considering that machine an insult to a true-blue hobbyist such as himself), he never made an *Electric Pencil *for the Commodore PET or Apple II. He also grew sick of the work of revising the software, never properly rewrote it to make it more modular, and eventually sold off the rights to others who failed to keep up with the competition.[13]

This left the field open for rivals. Commodore supplied its own word processing program, called simply Commodore Word Processor, at a price of $99.95. The most popular early word processor for the Apple II was created by John Draper, a.k.a. Captain Crunch, the infamous phone phreaker. While serving time at Alameda County Jail for hacking the phone system and programming at Apple during the day under work furlough, he began working on a compiler for the Forth programming language on the Apple II. Like Shrayer, he found the process of documenting his creation daunting, and so began writing a word processor to solve the problem. Released in 1979 by “Cap’n Software,” *EasyWriter *provided Apple II users with roughly equivalent capabilities to Electric Pencil, from which Draper borrowed liberally.[14]

*Word-Star *(later called WordStar)was another matter. In 1978, Seymour Rubenstein left his position as director of marketing at IMSAI to found MicroPro International, with the intention of selling serious business software for the microcomputer. He had no difficulty poaching IMSAI’s star programmer, Rob Barnaby, who saw the writing on the wall at IMSAI and already resented CEO Bill Millard for refusing to market his text editor, Next Editor or NED. The typical microcomputer code editor at the time was a *line editor *that allowed a user to create or modify one line of the program at a time. Barnaby created NED to make it possible to edit a whole screenful of code at a time. But it was not a word processor: it had no word wrap, text formatting, or print layout features.[15]

Rubenstein, however, unlike Shrayer, did careful market research on existing word processing systems and software and explicitly tasked Barnaby with making the best such package possible for Gary Kildall’s popular CP/M microcomputer operating system. For this serious software, Rubenstein intended to charge a serious price: $495 for the first edition of WordStar, released in 1979. Barnaby’s editor included many improvements over Electric Pencil. The software package had a modular design that greatly simplified software production and sales. MicroPro initially sold only one version of Word-Star, for Gary Kildall’s CP/M operating system (which abstracted away the user’s disk storage system). Then the user indicated at configuration time what type of display peripheral they used, and the software itself then patched in the appropriate subroutines for the user’s screen. Word-Star could render many types of formatting on-screen (Rubenstein’s later ads for WordStar popularized the phrase “What You See Is What You Get,” or WYSIWYG), featured pop-over text menus for special commands, and supported the ability to handle documents larger than the computer’s memory by moving data to and from disk as needed.[16]

I have not been able to find any screenshots of the very earliest versions of WordStar, but this screenshot of WordStar 3 from 1982 gives a sense of its greater sophistication, with a status bar and keyboard-navigable menus.

Indeed, the tightest constraint on doing real work on a microcomputer was access to enough memory to hold your digital artifacts. As one reviewer of Electric Pencil wrote, “[y]ou need at least 16K and either Level-I or Level-II [BASIC] to use the [Electric Pencil]. Memory is power, and if you’re going to do much writing you’ll want to obtain more memory. Since my system is 48K, I’ve been able to write some lengthy ditties without running out of memory too often.” Therefore, the rapid evolution of Moore’s Law was critical to the success of personal computer business software in the late 1970s: memory prices dropped by nearly a factor of four between 1977 and 1979 alone. The release of the Apple II Plus in June 1979 reflected this; if offered sixteen kilobytes standard for the same price as the original four-kilobyte Apple II.[17]

VisiCalc

Its inherently larger memory capacity than its rivals (up to forty-eight kilobytes without any auxiliary add-ons like the TRS-80’s clunky Expansion Interface) also helped make the Apple II attractive to the creators of the first VisiCalc.

Up to this point, personal computers were largely imitative: since the Altair and even before, the hobbyist dream had been to recreate the capabilities of a serious minicomputer in a price and form factor fitting to the budget and available space of an ordinary suburban home. In 1977, Apple, Commodore, and Tandy took the capabilities built up by hobbyists pursuing that goal over the previous several years (BASIC interpreters, video terminals, inexpensive dynamic RAM, etc.) and put them into an even more affordable and user-friendly package. In a slightly different twist on imitation, Wozniak also gave the Apple II the capability to play color video games because he wanted to recreate the experience of Atari arcade games like Pong and Breakout. However, nothing quite like *VisiCalc *existed prior to its appearancefor the Apple II in 1979.

That *VisiCalc *ended up as a personal computer program at all was an accident of timing and chance encounters. It began with Dan Bricklin. Bricklin studied computer science at MIT, and worked under Fernando Corbató (Corby) on the Multics time-sharing system. After he graduated in 1973, he got a job nearby at Digital Equipment Corporation (DEC). For the next several years he worked exclusively on business applications for manipulating digital text: first a typesetting system for newspapers, and then a minicomputer-based word processor, WPS-8. When DEC moved his group to New Hampshire he stayed in the Boston area, and, in 1977, returned to Cambridge to attend Harvard Business School.[18]

While at Harvard, Bricklin had the idea of creating a program that would allow real time, interactive computation, in the same way that word processing allowed real-time, interactive text editing. Rather than setting up input data, writing code, and then executing the code to process the input and generate output, all three steps would happen at once. He thought of this as three distinct “planes” of activity, and collapsed all three into a grid: a cell could contain an input, or it could contain a computation, in which case the interface would render the output in the cell and then the computation (or formula) in a separate status line (a concept borrowed from the Harris 2200 digital typesetter). The grid visualization came from the paper ledgers (sometimes called spreadsheets) that he was familiar with using for accounting in business school.[19]

His experiences at MIT and DEC put Bricklin at the cutting edge of interactive computing culture, but he was not a hobbyist and had no microcomputer of his own. He created his first prototype in BASIC on Harvard’s PDP-10. He wanted to turn it into a product, and the natural model was a self-contained number processing system, much like DEC’s WPS-8 word processor. For the hardware, he targeted the DEC PDT-11 Programmable Data Terminal, a $5000 smart terminal designed for compatibility with PDP-11 computer systems, and in September 1978 he began gathering materials and ringing up old contacts at Digital.[20]

At around the same time, however, Bricklin had a fateful conversation with a professor, who told him to talk to a recently-graduated Harvard business student, Dan Fylstra. We have already met Dan’s brother, Dave, author of the 1976 *BYTE *article “Homebrewery vs. the Software Priesthood.” Dan, an editor of BYTE, was even more deeply immersed in the hobby computer world, and by 1978 was one of the many new entrepreneurs looking to make a living (or more) by selling software for microcomputers. He saw his opportunity in publishing, not authoring his own software, and drew on his contacts in the industry to find marketable material for his new venture, Personal Software. His strongest seller was a game, *Microchess, *originally written by Canadian Peter Jennings for the MOS Technology KIM-1, in 1976 (it became a Personal Software property after Fylstra merged his company with Jennings’ Micro-Ware). Fylstra got distribution deals in hundreds of computer stores (and thousands of Radio Shack outlets, for the TRS-80 Microchess).[21]

A 1979 advertisement for Personal Software’s offerings. The superiority of the Apple II’s graphical capabilities (at right) is obvious compared to the simple tiles of the Commodore PET (center) and the chicken scratches of the TRS-80 (left). [BYTE, February 1979, 106]

After multiple conversations with Fylstra, Bricklin was convinced to publish through Personal Software, and pivoted to developing a new prototype in Apple II BASIC. There are both circumstantial and fundamental reasons that Fylstra and Bricklin chose the Apple II: circumstantial because it was the only machine Personal Software had available to loan; Fylstra was using his Commodore PET to continuously to duplicate software for publication, and his TRS-80 to run his accounting. Fundamental because both men appreciated Apple II’s superior technical capabilities (high-resolution graphics, up to forty-eight kilobytes of memory, and the newly-released Apple II Disk) and Apple Computer’s superior business strategy: better marketing and customer service, and open distribution with no lock-in to Radio Shack stores. Bricklin brought in Bob Frankston, an old friend and colleague going back to his Multics days as an MIT undergraduate, to write the final product in 6502 assembly language, with Bricklin’s guidance on product design. Due to the still-meager software authoring capabilities of microcomputers, Frankston used a university-owned time-sharing system develop the program, just as Bill Gates and Paul Allen had done with their BASIC four years before (though in this case it was MIT’s Multics, rather than Harvard’s DEC PDP-10).[22]

A screenshot of a later, MS-DOS version of VisiCalc. The status bar indicates which cell the user is currently viewing (B5), and the formula that calculates the visible value of that cell (B3-B4).

Fylstra released Bricklin and Frankston’s “problem-solving software” in the fall of 1979 as VisiCalc, for $99.95. In their first ad, they already recognized what made it special: it was a “what if?” machine:

Once your first projection is complete, you’re ready to use VisiCalc’s unique, powerful recalculation feature. It lets you ask “What if?” examining new options and planning for contingencies. “What if” sales drop 20 percent in March? Just type in the sales figure. VisiCalc instantly updates all other figures affected by March sales.

Or say you’re an engineer working on a design problem and are wondering “What if that oscillation were damped by another 10 percent?” Or you’re working on your family’s expenses and wonder “What will happen to our entertainment budget if the heating bill goes up 15 percent this winter?” VisiCalc responds instantly to show you all the consequences of any change.[23]

Many mainframe computer programs before had calculated values over a grid, but none had the interactivity of VisiCalc, updating the grid on the screen, in-place, after each change to the inputs or formulas. And, of course none of those programs were as accessible as VisiCalc, which could be picked up (along with the Apple II to run it) at your neighborhood computer shop. Financial professionals, business executives and small business owners all flocked to this new tool that let them quickly model problems, adjust parameters, and get concrete answers about the results, without guesswork. The term “spreadsheet” for this class of programs became popular only later in the 1980s; early articles and ads most often called VisiCalc a “worksheet.”[24]

VisiCalc demanded a powerful and expensive Apple II system, with thirty-two kilobytes of memory and a disk drive. But enough people were willing to pony up to make Personal Software’s name, quite literally: after sales of $3.7 million in 1980 and $14 million in 1981, the company re-branded as VisiCorp, with a whole suite of Visi- products. It also propelled Apple II to the front of the pack; after a slow start relative to Commodore and Radio Shack sales grew rapidly at the end of the decade, reaching 130,000 Apple IIs by September 1980 and fueling an explosive IPO three months later that made Jobs, Woz, and many other early employees into millionaires (at a time when that really meant something).[25]

Business Computing

By the time of VisiCalc, personal computing was unrecognizable compared to its state just after the Altair appeared in 1975. At that time, it had consisted of an eclectic group of hobbyists exchanging barely-functional mail-order machines and hand-typed newsletters. Five years later, 724,000 personal computers had been sold, with a total value of $1.8 billion. And they existed within a sophisticated, highly-competitive, and collectively very lucrative network of computer and peripheral manufacturers, software developers and publishers, full-color print magazines, distributors, retailers, and service vendors. Personal computers had entered the public consciousness, with new developments followed by outlets from the *New York Times *to Playboy.[26]

Meanwhile, VisiCalc, word processors, and other application software (such as accounting packages) had also created a market for *business *computers. These differed from hobby computers mainly in including more memory, storage and peripherals: business customers wanted a fully-functional ready-to-use machine without having to track down and figure out how to install various accessories and add-ons. Radio Shack quickly recognized the value of this market, and introduced the TRS-80 Model II in October 1978, with an integrated eight-inch disk drive and a base thirty-two kilobytes of RAM. But with a price starting at $3,450, it did not exactly fit within the cost-conscious Radio Shack brand, and sales never took off. The base model TRS-80 instead did double duty as a small-business workhorse, and its successor, the Model III, was explicitly marketed as such.[27]

A Radio Shack Model II ad, showing a fully configured, four-drive system at nearly $9,000. Note that it still lacks the production values of Apple ads from 1977. [BYTE (October 1979), 146]

VisiCalc, however, had made Apple II the business computer to beat, especially with the greater baseline memory capacity of the Apple II Plus. After a slow start relative to Commodore and Radio Shack, sales grew rapidly at the end of the decade, reaching 130,000 Apple IIs by September 1980 and fueling an explosive IPO three months later that made Jobs, Woz, and many other early employees into millionaires (at a time when that really meant something). A deal with Bill Millard’s rapidly growing ComputerLand stores gave Apple’s computers a comparable retail presence to Radio Shack’s. In March 1980, Microsoft’s SoftCard plug-in also allowed the Apple II to emulate a Z80 processor and run the complete CP/M software ecosystem, neutralizing Radio Shack’s advantage in that sphere. (Though by the end of that same year, sensitive ears might have heard rumors of the approaching storm that would flatten CP/M and nearly drown Apple).[28]

Commodore, meanwhile, never even tried to compete with its own business computer. Instead, their next product pivoted down-market, into the emerging class of “home computers”, a gentle euphemism for a programmable games machine. It is to the business of computer games that we will turn next.

---

[1] Robert L. Perry, *Owning Your Home Computer: The Complete Illustrated Guide *(New York: Everest House, 1980), 31.

[2] William J. Hawkins, “New Home Computers Can Change Your Lifestyle,”* Popular Science *(October 1977), 36.

[3] “Introducing Apple II,” *BYTE *(January 1978), 10; “Radio Shack TRS-80,” *BYTE *(October 1977), 43.

[4] Processor Technology, “Sol System Suggested Retail Price List,” June 1, 1977, https://www.sol20.org/articles/pricelist.pdf; Computer History Museum, “Oral History of Charles Ingerham ‘Chuck’ Peddle” (June 12, 2014), 63.

[5] Leslie Haddon, “The Home Computer: The Making of a Consumer Electronic,” Science as Culture 1,2 (1988)17-18; “Computer Will Save Fuel On G.M.’s ‘77 Toronado,” *New York Times *(August 10, 1976); Computer History Museum, “Oral History Panel on the Development and Promotion of the Intel 8008 Microprocessor” (September 21, 2006), 28.

[6] Charlotte Kent, “The Watch That Made Everything Now,” *Wired *(Nov 30, 2021) (https://www.wired.com/story/pulsar-digital-watch-design-time-history/); “The World’s First Mass Produced Computer,” *Digital Design *(October 1977), 86 (https://archive.org/details/pet-ad/mode/1up).

[7] David Bunnell, “The Last Word,” Personal Computing (September/October 1977), 128.

[8] Radio Shack, “Radio Shack TRS-80 Software Library” (1978) (https://archive.org/details/trs-80-software-library-1978); Richard N. Langlois, “External Economies and Economic Progress: The Case of the Microcomputer Industry*,” The Business History Review* 66, 01 (March 1992), 17, 41-50.

[9] Laine Nooney, *The Apple II Age: How the Computer Became Personal *(Chicago: University of Chicago Press, 2023), 88-91.

[10] Freiberger and Swaine, Fire in the Valley, 186-187. Shrayer’s only film credits on IMDB list him as a cameraman on two films by the French director Jean-Pierre Melville. In an interview he also indicated that he was once a cameraman on *Candid Camera *(Paul Freiberger, “Electric Pencil, First Micro Word Processor,” InfoWorld (May 10, 1982), 12). Little else about his biography seems to be concretely known. http://www.computer-timeline.com/timeline/michael-shrayer claims his dates as 1934-2006, but it’s unclear where this information comes from.

[11] “The Electric Pencil II,” *BYTE *(July 1978), 77.

[12] Joseph A. Greenleaf, “Michael Shrayer’s Electric Pencil,” *Personal Computing *(May 1979), 72-74; Michael Shrayer, “The Electric Pencil Word Processor Operator’s Manual” (1977) (https://www.sol20.org/manuals/pencil.pdf).

[13] Michael Shrayer, “Confessions of a Naked Programmer,” *Creative Computing *(November 1984), 130.

[14] Thomas J. Bergin, “The Origins of Word Processing Software for Personal Computers: 1976−1985,” *IEEE Annalys of the History of Computing *(October-December 2006), 35-36; “Word Processor,” COMPUTE! (Fall 1979), 14; Ken Silverman and John T. Draper, EasyWriter (Griffith, IN: Information Unlimited Software, 1979). (https://mirrors.apple2.org.za/ftp.apple.asimov.net/documentation/applications/misc/EasyWriter%20Manual.pdf).

[15] Jonathan Littman, *Once Upon a Time in ComputerLand, Revised and Expanded Edition *(New York: Touchstone, 1990), 144-146; Bergin “The Origins of Word Processing Software for Personal Computers,” 38.

[16] John C. Dvorak, “WordStar from MicroPro International,” *John C. Dvorak Software Review *(September/October 1979), 32. “Word-Star,” *BYTE *(November 1979), 64; Bergin, “The Origins of Word Processing Software for Personal Computers: 1976−1985,” 39.

[17] Joseph A. Greenleaf, “Michael Shrayer’s Electric Pencil,” *Personal Computing *(May 1979), 72; “Historical price of computer memory and storage,” *Our World In Data *(https://ourworldindata.org/grapher/historical-cost-of-computer-memory-and-storage).

[18] Martin Campbell-Kelly and Paul Ceruzzi, “An Interview with Dan Bricklin and Bob Frankston,” Charles Babbage Institute (May 7, 2004), 5-11.

[19] Campbell-Kelly and Ceruzzi, “An Interview with Dan Bricklin and Bob Frankston,” 12-13.

[20] Campbell-Kelly and Ceruzzi, “An Interview with Dan Bricklin and Bob Frankston,” 13-14; DataPro Research Corporation, “Digital Equipment Corporation PDT-11 Terminal Family,” August 1979 (http://www.bitsavers.org/pdf/datapro/programmable_terminals/Datapro_C21_Digital.pdf).

[21] Thomas Haigh, “Oral History of Dan Fylstra,” Computer History Museum (May 7, 2004), 15-16; Nooney, The Apple II Age, 90-91.

[22] Haigh, “Oral History of Dan Fylstra,” 14-15; Campbell-Kelly and Ceruzzi, “An Interview with Dan Bricklin and Bob Frankston, 17, 19.

[23] Personal Software, “Solve Your Personal Energy Crisis. Let VisiCalc Power Do the Work,” *Personal Computing *(September 1979), 1.

[24] Steven Levy, “A Spreadsheet Way of Knowledge,” *Wired (October 24, 2014 [November 1984]) (https://www.wired.com/2014/10/a-spreadsheet-way-of-knowledge); “VisiCalc,” Creative Computing *(November 1980), 19. An interactive UI like VisiCalc’swas theoretically possible on a mainframe with a smart terminal incorporating a screen buffer. It’s also again worth noting that *VisiCalc *lacked many features and algorithmic tricks that newer spreadsheets incorporate. For example, *VisicCalc *recalculated cells according to a strict order (e.g. A1, A2, A3), without any support for forward dependencies (e.g. A1 = A2 + A3). Dan Fylstra and Bill Kling, VisiCalc: User’s Guide

*for the ATARI 800 32K *(VisiCorp, 1981), 70-71.

[25] Richard P. Rumelt, “Visicorp 1978-1984 (Revised),” (2003), 4, 8 (https://web.archive.org/web/20031101141127/https://www.anderson.ucla.edu/faculty/dick.rumelt/Docs/Cases/Visicorp.pdf).

[26] “Machine of the Year 1982: The Computer Moves In,” *Time *(October 5, 1983); Peter J. Schuyten “Commodore Prepares Challenge To U.S. Home-Computer Giants,” *New York Times *(October 1, 1980); Phil Wiswell, “The Friendly Home Computer,” *Playboy Guide: Electronic Entertainment *(Fall 1980), 113-114.

[27] Welsh and Welsh, Priming the Pump, 42; Ira Goldklang, “TRS-80 Computers: TRS-80 Model III,” (https://www.trs-80.com/wordpress/models/model-3).

[28] Michael Moritz, *Return to the Little Kingdom *(New York: Overlook Press, 2009), 243, 276-277; Bagnall, Commodore, 136-137; David Bunnell, “A PC Exclusive Interview with Software Guru Bill Gates,” *PC *(Feb-Mar 1982), 20.