Computer Chronicles Revisited 85 — WordStar 4.0, WordPerfect 4.2, Microsoft Word for MS-DOS 4.0, and OfficeWriter 5.0

The history of computer word processing applications can be divided into three main periods, each tied to a specific platform transition. In the first period–the late 1970s and early 1980s when CP/M machines dominated the microcomputer market–WordStar was the gold standard. After the IBM PC came along and MS-DOS displaced CP/M, WordPerfect similarly overtook WordStar. Then, as Windows was finally accepted by the masses at the start of the 1990s, Microsoft’s Word usurped WordPerfect as the one word processor to rule them all. (And if we extend things to the present, in the post-PC age there is effectively a commercial word processing duopoly with Microsoft 365 and Google Workspace.)

This next Computer Chronicles episode from March 1987 aired during this second period when WordPerfect had taken the lead. WordStar was struggling to regain its footing after a number of stumbles. And Word was still thought of primarily as a Macintosh application. There were also dozens of smaller companies still competing for a dwindling share of the market.

This episode was actually the first in a four-part series on “business applications.” The next two programs focused on spreadsheets, while the final delved into database management. To begin this word processing episode, Stewart Cheifet showed co-host Gary Kildall what might be dubbed the original word processor–a pencil. Cheifet quipped that early word processors didn’t do much more than the simple pencil. But they were much more sophisticated these days. And some people were saying that word processors were now evolving into desktop publishers.

Kildall agreed that desktop publishing was an evolution from word processors. But it was a lot more. Desktop publishing involved the integration of text and graphics, taking into account the structure of the document and the page layout. So desktop publishing was still far more complicated, and he thought the two applications would continue to remain separate for some time.

Single-Programmer Shareware Competed Against Big Corporate Word Processors

Wendy Woods presented her first remote segment, which focused on a low-cost word processing program called PC-Write. Woods said that most software packages came with a stern warning about making more than your fair share of copies. So why did PC-Write ask its owners to make as many copies as they liked? Because PC-Write belonged to a class of programs called shareware, and its success depended on as many people as possible trying it out for themselves.

Paul Berry, a professional writer, told Woods that the nice thing about PC-Write was that you could try it out and say, “Yes, this is what I want,” or “No, this isn’t what I want.”

Woods said that if you couldn’t get a copy from a friend, PC-Write diskettes were available from the publisher for $16. If you decided you liked the program, you sent in the $89 registration fee in exchange for a manual and updates. Woods added that PC-Write was not a “bare-bones” program. It included a spelling checker, mail merge, and laser printer support. The program also allowed for some split-screen editing and custom character sets.

Jay Siegel, a writer and programmer, told Woods that PC-Write was written in the Pascal programming language, and he thought it was very fast and efficient because it was done by a single programmer. In his experience, when you found a program that was a pleasure to use, it usually had just one programmer behind it.

Woods said the idea of shareware seemed far removed from the corporate mainstream of commercial software, which could be explained in part by its unusual origins. Siegel said that shareware had its roots in ham radio. The original people who developed the shareware concept came from the amateur radio community, which was based on the idea that the airwaves were free and the goal was an exchange of information.

Keeping Power Users Happy

Walter Feigenson and Dan Lunt joined Cheifet and Kildall for the first studio segment. Feigenson was vice president for product marketing with MicroPro International, the company that developed WordStar. Lunt was vice president of marketing with WordPerfect Corporation.

Kildall opened by asking Feigenson what kind of features you needed to have in a word processor these days to keep it on top. Feigenson said the primary method was listening to the users. He noted that users of word processors were especially vocal. They loved to give you suggestions, and all you had to do was listen to them.

Lunt added that as WordPerfect listened to its users, a large number of them were attorneys, accountants, and professional people, and they all wanted features specifically directed at their industry. For example, WordPerfect received a lot of requests for line-numbering and “table of authorities” features from attorneys, so they were added.

Kildall turned back to Feigenson and asked for a demonstration of WordStar 4.0. Feigenson opened a document and noted that the new version retained the “classic look” of the traditional WordStar interface while adding 125 new features. One of those features was the ability to point to a filename to open it, something that many people had asked for over the years.

Feigenson started editing the sample document. Demonstrating another new feature, he executed a macro to draw a box around the title of the document. (This could also be done manually.) There was also a faster in-context spelling checker. The program flagged a word it didn’t understand–the name of a city–and suggested alternative spellings. The user could select an alternative by typing the number next to it. Feigenson said the spell checker could also understand words spelled phonetically. So if you typed “newmonya,” the spelling checker would suggest “pneumonia.”

WordStar also came with what Feigenson described as a “very elegant” thesaurus. By positioning the cursor over a word and pressing a hotkey, the software displayed a list of suggestions. You could also page through the thesaurus database to look for a specific word.

While Lunt setup his own demonstration, Cheifet asked Feigenson about the biggest complaint people had about the previous versions of WordStar. Feigenson said WordStar hadn’t changed for a long time, but the technology had changed since the last update. And there were hundreds of things that people were asking for. So it wasn’t hard to figure out what people wanted, such as an undelete feature and a thesaurus.

Cheifet noted that WordPerfect had recently released a new version that had “risen to the top of the pile” in the word processor market. He asked Lunt to demonstrate some of those new features in WordPerfect 4.2. Lunt said that one of the goals when designing WordPerfect initially was to make the word processor look as much like a typewriter as possible. So when the screen for WordPerfect came up, you would see what appeared to be a blank sheet of paper. There was nothing that you needed to go through before that.

Lunt said that WordPerfect emphasized a simple interface with a single-keystroke approach for features. So if you were typing text and wanted to make something bold, you simply pressed the “bold” key. If you wanted underlining to appear, you pressed the “underline” key. (In this version of WordPerfect, formatting was indicated by color, so bold text appeared in blue, while underlined text appeared in yellow.) He also showed how you could center text and apply automatic text-wrapping to the end of a line.

Cheifet asked about WordPerfect’s approach to the thesaurus. Lunt said all the user had to do was point to a word in a document, or type a word, and press the “thesaurus” key. This brought up a list of both suggested synonyms and antonyms.

Cheifet asked about working in columns, which was a feature that WordPerfect touted. How easy was it to do columns? Lunt said it was very easy. He pulled up another sample document–the text of the United States Constitution–to demonstrate. The document was set to three “newspaper-style” columns, i.e., where text from column one snaked to column two and then on to column three. WordPerfect handled all of this formatting automatically. So if you added text into the document the software would adjust the columns accordingly.

Kildall asked to what extent these word processors could handle graphics. Lunt said WordPerfect had “very limited” graphics capability. The program could handle graphs in a character format. It could also block a space within a document for another program to insert a graphic. But in the future, WordPerfect’s goal was to make it possible to mix graphics and text together just using its software.

Feigenson said WordStar was currently in a similar place. You could insert character-based graphics only. And there were 10 function keys allocated to drawing lines and boxes using characters. You could also use a third-party program to add graphics to a WordStar document.

Competitors Emphasized Macro-Based Features

Lewis Levin and Jim Kesler joined Cheifet and Kildall for the next segment. Levin was the product manager for Microsoft Corporation’s Word. Kesler was the product manager for OfficeWriter, a word processor published by Office Solutions, Inc.

Kildall noted that a good word processor today needed to have certain basic features, such as a spelling checker and hyphenation. But what else did you need to have a quality product in the current market? Levin said Microsoft’s philosophy was that the word processor should help the user as they wrote. The word processor should help you organize your document and help you figure out what you were trying to say. It was also important for the word processor to show you what the document would look like when you printed it out, especially when using a modern laser printer.

Kesler said that a good word processor had to be a “diverse” product. It needed all the features that power users demanded. Yet it still needed to be easy enough for a broad range of users to implement.

Kildall asked Levin for a demonstration of Microsoft Word for MS-DOS. (This was version 4.0.) Levin showed off the outlining feature of Word, which he said distinguished the product from its competition. He pulled up a sample document–a product brochure–and showed the “outline view.” If he wanted to look at the details of any one section, he could expand or collapse the outline accordingly. The outline was fully integrated into the word processor. This made it helpful when trying to reorganize big sections, because you could pick up one of the points on the outline and move it, which took along all of the accompanying text.

Levin also demonstrated the on-screen formatting features of Word. Part of the title was in italics. (As opposed to the earlier WordPerfect demo, Word actually displayed italics rather than representing them by using a different color.) Levin pulled up a list of different character formats that Word could display, such as double underlining and strikethrough. Word also enabled you to combine various formatting with a single macro key using what he called style sheets. He added that Word supported a number of PostScript fonts for the Apple LaserWriter printer. Kildall pointed out that unlike the other word processors demonstrated so far, Word used a bitmap in order to display the document formatting on-screen.

Cheifet asked Kesler how his product, OfficeWriter, could compete effectively against the more-established products like WordPerfect and Word. Kesler said OfficeWriter had seen the most success with Fortune 500 companies, and recently they had started focusing on resellers as well.

Kesler then provided a demonstration of OfficeWriter. The opening screen displayed a menu. Using the function keys, the user could select an action, such as creating or printing a document. Kesler showed how OfficeWriter could automatically draw lines or boxes around text–say to create an organizational chart–without the user having to create each one individually. Cheifet noted this didn’t require the use of a mouse. Kesler said yes, you could simply use the numeric keypad. Kildall emphasized this particular demo was accomplished using OfficeWriter’s built-in macros.

Kildall asked about the “boilerplating” feature in OfficeWriter. Kesler said it was similar to marcos. Boilerplating allowed users to incorporate paragraph phrases and statements from other programs and bring them into your document. He provided a quick demonstration. Essentially, the boilerplate text was tied to a macro key.

Cheifet asked Kesler if there was any particular applications that OfficeWriter was unique for. Kesler said the product’s emphasis had always been on connectivity and compatibility. OfficeWriter particularly shined when it came to graphics integration and desktop capabilities. The software was logical and intuitive to use yet still offered the features that power users demanded.

Word on the Mac Inched Closer to PageMaker

Wendy Woods returned for her second remote segment, reporting from TechArt, a San Francisco-based graphics design firm. (This was TechArt’s second appearance; Woods previously featured the company in a 1985 episode on laser printers.) Woods noted that TechArt had been using Microsoft Word 3.0 for the Macintosh for several months. The program had received “rave reviews” from TechArt’s staff, who depended on good word processing programs to create professional documents.

Diane Burns, TechArt’s president, told Woods that the single feature most in demand that had been added to Word was a spelling checker. Word already had a spell checker on the IBM PC version. And now a version of it was brought over to the Macintosh in version 3.

Another important feature was the addition of style sheets, which Levin discussed in the prior segment on the PC version. Burns said the style sheets let her assign a complex format “spec” to a small code, and then attach that code to subsequent paragraphs so they would take on that formatting.

Woods said Word 3.0 was also exciting because it had elements found in a desktop publishing program. Graphics could be inserted, and page layout could even be previewed. Burns said she could place a graphic side-by-side with the paragraph that she selected. She showed a sample page layout (or print preview) screen. It wasn’t a replacement for Aldus’ PageMaker, as you could not edit the document in this mode, but it still provided the ability to do something that wasn’t possible before.

The Future of Word Processors

For the final segment, Michael Miller and Eric Alderman joined Cheifet and Kildall. Miller was executive editor at InfoWorld. Alderman was a columnist with Computer Currents. (Miller previously appeared in a May 1985 Computer Chronicles episode to discuss the Japanese computer industry.)

Kildall asked Miller how the marketplace was split up right now for word processors. Miller said the business marketplace was divided into three categories:

Miller added that a lot of these products had overlap. That is, they could read files created by other word processors.

Kildall asked Alderman if he saw word processors eventually morphing into desktop publishing programs. Alderman said he thought that they would. Word processing programs would eventually take on a lot of the more advanced features that we saw earlier in the program, such as Microsoft Word’s style sheets. User customization and laser printer support would also become important going forward.

Cheifet asked Miller and Alderman–both professional writers–which word processors they used. Miller said he split his time between Word, WordPerfect, and WordStar, running all three at different points. Alderman said he used Word and WordPerfect.

Apple Promotes IIgs as Rumors Swirl of New 32-Bit Macintosh OS

Stewart Cheifet presented this week’s “Random Access,” which was recorded in March 1987.

Venture Capitalists Quickly Ousted OfficeWriter Founders

Since WordPerfect and Microsoft Word will continue to battle for word processor dominance over the next few years in our Chronicles timeline, I’ll focus the remainder of this post on the two also-rans, WordStar and OfficeWriter. I’ll start with OfficeWriter since that is the simpler story.

Jim and Mary Crist, a husband and wife from Madison, Wisconsin, founded Office Solutions, Inc., the company behind OfficeWriter. In late 1982, Jim Crist worked in technical support for Wang Laboratories, while Mary was a computer operator and programmer. In late 1982, they decided to develop their own word processing program for the IBM PC. Working nights and weekends for several months, they managed to create a working prototype by January 1983.

After demonstrating their word processor to local computer users in Madison, the Crists incorporated Office Solutions in April 1983. Jim Crist quit his job at Wang to focus on the new company as vice president of marketing. Mary Crist continued working as a consultant while also serving as vice president of finance. They brought in Jan Eddy to run the company as president. Eddy previously worked in data processing for Land’s End, the Wisconsin-based clothing retailer.

OfficeWriter proved successful out of the gate. A 1986 article in Madison’s Capital Times said Office Solutions reported $2.2 million in annual sales–a 30-fold increase over the first three years of operations. The company now employed 42 people and was on track to reach $4 million in sales by the end of the 1987 fiscal year.

Naturally, the success attracted venture capital. Eddy sought funding early in her tenure as president, which meant diluting the control that she and the Crists held in terms of stock and board seats. And while the company grew rapidly, it apparently wasn’t rapid enough. Eddy later recalled that the directors representing the VCs created “financing and operational roadblocks,” which prompted the Crists to look for a buyer.

In August 1988–just a couple of weeks after moving into a new 31,000 square-foot headquarters in Madison–Office Solutions was acquired by Software Publishing Corporation (SPC), a California-based company that published the word processing program Professional Write. SPC reorganized Office Solutions as Software Publishing Wisconsin. Eddy stayed on as general manager for a time but ended up leaving to join another husband-and-wife startup venture, Wingra Technologies.

For the next few years, the former Office Solutions continued to grow as Software Publishing Wisconsin. A June 1991 report in the Wisconsin State Journal noted the subsidiary now had 112 employees and reported annual sales of $30 million, which included both OfficeWriter and Professional Write Plus. Altogether, the Wisconsin office generated about 20 percent of Software Publishing’s total revenues.

Alas, the good times didn’t last. In 1994, SPC decided to close the Wisconsin office as part of a reduction of roughly half the company’s workforce. By this point there were just 65 people employed in Madison. About half of them were reportedly transferred to another company that purchased Software Publishing Wisconsin’s customer support operations. SPC itself didn’t last much longer and was sold in 1996 to New Jersey-based Allegro New Media, Inc. (later renamed Vizacom).

WordStar’s Founder Feuded with Venture Capitalist Backer

The history of WordStar began with Seymour I. Rubinstein. In the 1970s, Rubinstein served as director of marketing for IMSAI Manufacturing Corporation, which developed one of the earliest microcomputers, the IMSAI 8080. According to Rubinstein, he signed the first contract with Gary Kildall to bring the latter’s CP/M operating system to the IMSAI machine.

In mid-1978, Rubinstein left IMSAI and together with a former colleague, John R. Barnaby, founded MicroPro International. Rubinstein said MicroPro’s goal from the outset was to produce a word processing program for CP/M computers that could rival the standalone word processors sold by companies like IBM and Xerox. Those machines often cost upward of $15,000, so it made sense to develop a word processing application for the less-expensive micros.

While working at IMSAI, Barnaby developed an ASCII text editor called NED, which was designed as replacement for CP/M’s ED. Barnaby used the knowledge he gained from creating that program to develop WordMaster, MicroPro’s first word processing program. Together with a data sorting utility, aptly named SuperSort, Rubinstein debuted both programs at a September 1978 computer show in New York City, according to John C. Dvorak. Boosted by $12,000 in sales from that show, Barnaby and another programmer, Jim Fox, spent the next four months developing a more powerful word processing program–WordStar–which only kept about 10 percent of the WordMaster code. MicroPro formally debuted WordStar for CP/M at Jim Warren’s 1979 West Coast Computer Faire.

WordStar was an immediate hit upon its 1979 release, generating $500,000 in sales during its first year on the market. By 1981, that figure jumped to more than $5 million. In March 1984, MicroPro International went public, raising $23.1 million in its initial offering.

Unfortunately, about two months before the IPO, Rubinstein suffered a serious heart attack. His health recovered, but Rubinstein later said that the key venture capitalist who had backed MicroPro, Fred Adler, used the medical emergency as a pretext to wrest control of the company from him. Rubinstein said that one of Adler’s employees, H. Glen Haney, came to him in the hospital and demanded he sign documents converting Rubinstein’s stock into non-voting shares. Otherwise, Adler would call off the IPO. Rubinstein said he signed under duress and resigned as CEO soon thereafter. Rubinstein remained chairman but Adler installed Haney as the new president and CEO.

Perhaps in an effort by Adler to do some damage control, a 1985 Los Angeles Times report suggested Rubinstein’s ouster was justified by his management of MicroPro International during its early period of rapid growth:

“Everything that made him a great entrepreneur made him a failure as a manager,” said [former MicroPro marketing director Will] Luden, who describes himself as an admirer of his former boss.

The company continued to hire “inspired amateurs” as it grew, he said, rather than the business specialists that it needed. Budgeting was inadequate, so that when the company began losing money in 1982, officials were not sure whether it was from overstaffing or because other costs were too high, Luden said.

“We’d slap a hiring freeze, or go on a cost-cutting campaign, without really knowing,” said Luden. “We didn’t know where the hemorrhage was, so we slapped a bandage across the whole body.”

Now as we’ll see, Adler appeared to be just as much in favor of this managerial chaos as Rubinstein–in fact, he was almost certainly behind it–but there was no doubt that MicroPro’s periodic cost-cutting campaigns proved more harmful than helpful. Indeed, one particular round of employee cuts ended up creating a major problem for MicroPro and WordStar in the form of a new competitor. In 1982, MicroPro laid of three of its programmers–Peter Mierau, Stan Reynolds, and Richard Post–shortly after Mireau completed a hasty rebuild of Barnaby and Fox’s original CP/M codebase for WordStar, which was somehow lost. (Barnaby had left the company in late 1979.) Stung by the disloyalty, the three men decided to create their own company to produce a WordStar clone.

The programmers quickly found an ally in future Computer Chronicles contributor George Morrow, who agreed to invest $100,000 in capital in the new venture, which was named NewStar. This wasn’t an act of altruism on Morrow’s part. Like a lot of CP/M computer makers, Morrow Designs often sold their machines pre-bundled with application software. Morrow had been bundling his machines with WordStar. So by investing in this new clone venture–in exchange for a free license–Morrow stood to save quite a bit of money for his own business.

In September 1983, NewStar released NewWord, a word processor that mimicked about 80 percent of WordStar’s functionality. This was quickly followed up in 1984 with a 2.0 NewWord release that added a number of features that were missing from WordStar, such as an undelete function, a spelling checker, and laser printer support.

Meanwhile, MicroPro struggled to update its aging WordStar. Before his ouster, Rubinstein decided to purchase a UNIX-based WordStar clone developed by Edward de Jong, which was used as the base for a completely revamped WordStar. With de Jong as the lead programmer, MicroPro released this new WordStar 2000 in 1985.

To be clear, WordStar and WordStar 2000 had little in common other than the fact they were both word processing programs published by MicroPro. WordStar 2000 used a different file format and a more simplified command structure than classic WordStar. Of course, WordStar 2000 also added a number of features missing from the older application. MicroPro also released an even more simplified word processor called MicroPro Easy to cover the lower end of the market.

The problem was that many of the new features in WordStar 2000 were also available in NewWord–and existing WordStar users didn’t have to learn a new interface. Not surprisingly, many legacy WordStar users moved over to NewWord. At the same time, users looking for a simpler interface not burdened by 1970s programming standards flocked to WordPerfect.

Indeed, WordPerfect executives were just as befuddled by MicroPro’s stumbles with WordStar 2000 as everyone else. W.E. “Pete” Peterson, then WordPerfect Corporation’s executive vice president, wrote in his memoirs that sales of WordPerfect 4.0 benefited immensely from MicroPro’s decision to destroy its credibility with its most loyal customers:

[W]ord processing users had a religious-like zeal for their favorite products. For most people, switching from one product to another was almost unthinkable. WordStar was especially hard to learn and master, and fans of the product defended it with an irrational fervor. Their loyalty was similar to that of a mother who has given birth to a very ugly baby. It was almost impossible to get expert WordStar users to admit their product had any flaws.

The diehard WordStar users pushed their word processor on unsuspecting friends and co-workers, and this very vocal group represented our biggest obstacle to convincing new customers to buy our product. Luckily, Micropro made our job easy by attempting to replace WordStar with WordStar 2000. Micropro tried to fix all the problems in the original WordStar, and rather than making their loyalists happy, they alienated them. Their strongest supporters had bonded to the program’s quirks and problems and had come to believe that the difficulties were desirable. Worst of all, WordStar 2000 was very slow compared to the old WordStar. Even a mother could not have loved the new product.

During this crucial 1985-1986 period, WordPerfect started to overtake WordStar and never looked back. MicroPro panicked. Well, Fred Adler panicked. He ousted Haney, his handpicked CEO, in August 1986 and brought in an industry outsider, Leon Williams, to take over as president. Adler also laid off about 20 percent of the MicroPro staff for good measure.

Williams’ first priority was to try and regain the trust of the WordStar customers burned by the WordStar 2000 debacle. To accomplish that, he ended up buying MicroPro’s main competitor for those same customers–NewStar. Walter Feigenson, who had joined NewStar as its vice president of marketing, helped negotiate the sale of the company to MicroPro for $3 million, which was announced in October 1986. As part of the deal, Feigenson and several key NewStar employees, including co-founder Peter Mierau, received one-year contracts to stay on at MicroPro.

WordStar 4.0, the product that Feigenson demonstrated in this Chronicles episode, was effectively the final NewWord 3.0 code combined with some elements from the last release of the classic WordStar. The new-old Wordstar proved successful enough to put MicroPro back into the black. The company reported a $2.9 million profit in 1987 after sustaining a $1.2 million loss in 1986.

But in 1988, MicroPro slipped back into the red with a $5.1 million loss. Fred Adler reacted as you might expect. He fired Leon Williams in October 1988 and announced another round of 20-percent layoffs that December. Adler promoted chief operating officer Gari Grimm to take over as president and CEO. One of Grimm’s first moves was to change the company’s name from MicroPro International to WordStar International.

Grimm didn’t last long either. She resigned in October 1990 after WordStar reported yet another quarterly loss. The company continued to limp along until February 1994, when WordStar disappeared in a bizarre three-way merger with Spinnaker Software Corporation and SoftKey Software Products Inc. The story of that deal will almost certainly be the subject of a future post.

Notes from the Random Access File