The Story of Tandy – 10 years after the TRS-80 Model 1

Tandy-Story-10years

This article was first published on 80 Micro magazine, issue August 1987 and reproduced here with the “blessing” of its author, Ron White. The article was revised and illustrated to offer an interesting perspective on how Tandy went from that point in time until today. The story was chosen as the best feature among all the CPM publications that year.

Ron White is the former executive editor of PC/Computing magazine and the author of the best-selling How Computers Work 10th Edition The Evolution of Technology. The book’s Facebook page is at http://tinyurl.com/HCW10th

—–

It all started almost 40 years ago in a converted used-car showroom…

tandy_logo

 

By any logical assessment back in the 1970s, no one should have built a microcomputer to sell to a mass market. A few hobbyists were building computer kits that, more than anything else, tested their builders’ endurance and patience. They certainly didn’t do anything as useful as balancing a checkbook or even playing Space Invaders. The typical home consumer and small business weren’t shouting for their own computers. If most people thought about computers at all, it was as hulking giants that occupied air-conditioned rooms and were attended by white-robed priests who spoke In arcane terms about “kilobytes”, “BIOS,” and “dynamic RAM.” Still, some thought there might be something to these new machines.

These people were the hardware hackers, mostly on the West Coast, who pursued personal dreams in garages and one-room rented offices. The established computer companies—IBM. Hewlett-Packard, Digital—weren’t interested in downscaling their big machines into hobby-shop toys. But one established company thought there might be a future in building and selling a small computer. It wasn’t a computer company. It was Radio Shack, located in Texas, equally far from either of the hotbeds of computer innovation surrounding San Francisco and Boston. Radio Shack and its parent company. Tandy had never built a computer before. Their specialty was electronic parts for do-it-yourselfers and low-cost consumer electronics. If anyone had thought about it back then, Radio Shack might have seemed an unlikely candidate to help launch the American microcomputer revolution.

But some people at Tandy did think about it and decided that microcomputers might be the latest fad—like the CB radio market that had just boomed for a year or so and then quickly went bust. The executives at Tandy figured it was worth a gamble— a small gamble. They never suspected that microcomputers would go on to change the structure of American business and that Tandy would become a major force in that change.

Abhorring a Vacuum

No one in the twin granite towers of Tandy’s Fort Worth headquarters remembers the exact moment someone there conceived the idea of building a microcomputer. But in 1976, Tandy was looking for a market to exploit. The bottom had dropped out of CB radios, which had been a big moneymaker for Radio Shack stores. The loss of CB sales was not going to bankrupt Tandy with its diversified product line. But the loss left a vacuum in Tandy sales.

cb_23-channel

At the same time, some of Tandy’s electrical engineers were ordering something called the MITS Altair 8800 from an obscure company in New Mexico. The Altair was something only a hobbyist could love; it had no keyboard and no video display, and it had absolutely no practical use. But a lot of people were buying it, and a lot of the electronics magazines were writ- ing about it.

John Roach, a Tandy executive, read the articles and noticed the company’s engineers talking enthusiastically about their new toys. Roach thought Tandy should at least be looking into the new fad. A no-nonsense Texan with math, physics, and MBA degrees from Texas Christian University, Roach’s background was in computers. He had managed the data- processing department for Tandy before becoming its vice president of manufacturing. It was experienced that would eventually prove valuable in the development of Tandy computers, which in turn would help boost Roach to chairman of the board and chief executive officer at Tandy.

The engineering and merchandising departments got together to design a product that might appeal to the same people who were buying the Altair. The engineers came up with a plan for a computer kit that would outdo the Altair: it would have a keyboard and a monitor—more or less a real, functioning computer instead of an experiment for electronics freaks. Al- though the kit idea more than matched the marketing of the Altair, it was nixed by BernSe Appel. then vice president of merchandising and now president of Tandy’s Radio Shack retail division. Appel decided that even Radio Shack’s staple customers, who were no strangers to the soldering iron, could never put together something as complex as a computer.

“We decided there were too many problems likely to be encountered by the customers,” Roach recalls. “The chips were sensitive to electrostatic discharges, and it was actually harder to test the parts unassembled than it was assembled.”

A kit would also require extensive, lay- man-language Instruction manuals and would create nightmares for Radio Shack’s service department. So the company decided to do what, until then, no company had ever done before—sell a to- tally assembled microcomputer, complete with keyboard, monitor, processor, memory, and a programming language. It was not something any company, even one with the extensive in-house manufacturing experience that Tandy had, could whip out quickly. That’s where Steve Leininger came In.

Something Was Brewing

Leininger was a member of the Homebrew Computer Club, the loose hodgepodge of computer hackers located in what was coming to be called Silicon Valley. The club was where another electronics nut, Steve Wozniak, was showing off a rudimentary circuit board that would eventually become the Apple computer and Radio Shack’s most serious rival. Leininger, like many other members of Homebrew, had de- signed and built his own computer—two or three of them, in fact—and he attended the twice-monthly meetings of the club, where the concepts of proprietary information and making a fortune did not yet exist. Instead, something called the “Hacker Ethic” reigned. Members freely exchanged ideas and designs and occasionally “liberated” software or even microchips from companies with reactionary ideas a
bout property rights.

In contrast to other members of Homebrew, Leininger was more conservative, meaning that he bathed regularly, got haircuts at least semi-regularly, and was married to a flesh-and-blood woman instead of a growing conglomeration of wiring and electronic components that other members spent their nights with Leininger even had a regular job, designing chips for National Semiconductor, That’s where several executives from Tandy, led by Don French, met Leininger as they were on a fishing expedition to electronics manufacturers to find new technology that they might use in Radio Shack products.

Leininger was introduced to the Tandy representatives as someone working on a tiny version of Basic to be incorporated into a cheap microprocessor. While they were talking, the Tandy visitors mentioned that they’d like to see one of the California computer shops they’d heard about, Leininger gave them directions to the Byte Shop in Santa Clara, the second store to open in a chain of shops that specialized in selling computer components to hardware hackers. Later that day when the Tandy executives visited the shop, the person greeting them from behind the counter was Leininger, making a little extra money moonlighting.

After the executives were back in Fort Worth and the discussion turned to hiring someone to design a microcomputer.

French remembered this kid in California. Not only was he versed in programming and hardware, he had experience in the soul of Tandy operations, electronics re- tailing. He seemed like a good candidate for the Job.

The call Leininger received from Tandy was well-timed. He had been unhappy at National Semiconductor because he had been passed over for stock options. Leininger’s wife, to whom he had been married for six months, had a master’s degree in geology, but the best job she could find was the breakfast shift at McDonald’s. Tandy wasn’t offering much more money than Leininger was making at National Semiconductor, but the fact that Texas has no state income tax and that they would be in the middle of oil country made Leininger accept the job.

“Almost the universal reaction from my family and friends was, ‘You’re going to work for Radio ShacW ” he remembers. “Radio Shack had this hyper- schlock image—you know, buying out-of- spec parts, that sort of thing. I think they finally woke up to that about 10-12 years ago, but It’s taken a lot of work to get to where they are now—almost respectable.”

The Department That Wasn’t

When Leininger arrived in Fort Worth in 1976, he discovered the company wasn’t set up to build microcomputers. “There was no department for building computers,” says Bill Schroeder, who later helped create one of the operating systems for the new machine, “There was no one to manage the manufacturing. There was no lead time for the acquiring of software products— none of this stuff.” The lack of any master plan for going into the computer business would have surprised no one familiar with Tandy in those days. It was still run in an informal seat-of-the-pants style that had characterized it since 1918 when Dave Tandy and his partner, Norton Hinckley, bought a supply of shoe leather for resale to cobblers throughout Texas. When Tandy’s son, Charles, took over the business, he created a couple of retail outlets to sell supplies to leather-crafting hobbyists of the type he had run into while serving in the Navy in World War II.

The shops were a success and led to more outlets. In 1952, Tandy acquired a failing manufacturing plant in New England. The acquisition established a pattern for all Tandy operations, including its future venture into computing: The only products carried in its stores would be those with house labels on them, either made in Tandy’s own plants or made by other firms exclusively for Tandy.

By 1961, Tandy had 125 stores. The following year Charles Tandy came across Radio Shack, a chain of nine retail stores and a mail-order company in the Boston area that catered to ham-radio operators. He quickly whipped up a deal that in three years gave Tandy 85 percent ownership of Radio Shack. At the same time, it turned Radio Shack’s $4 million of red ink into profit with $20 million a year in sales. The stores by then represented 40 percent of Tandy’s income.

radioshack-600x371-600x3711

The number of Radio Shack stores continued to grow. By the time Tandy began thinking seriously about building com- puters, it had thousands of Radio Shack outlets worldwide generating pre-tax prof- its of over 40 percent. While hobbyists continued to prowl its ever-present bins of electronic parts, the stores had diversified into consumer electronics: TVs, stereos, games, and novelties.

For all its big-time success, however, there was a conspicuous lack of formal planning. A typical marketing strategy didn’t involve expensive studies that spilled out a lot of gibberish about demographics, product positioning, or image building. Instead, the stores simply stocked a minimum inventory of a new product. If it sold well, the stores increased the inventory; if the product was a dud, the loss was bearable. This strategy has also been used to promote a new cheap erectile dysfunction drug. The strategy has caused Radio Shack to miss some opportunities. During the Christmas season of 1983, Radio Shack stores ran out of video cassette recorders because management wanted to keep inventories low until they saw if VCRs, then only two years old, would catch on. It’s a safe, conservative strategy, and if it means that at times Tandy missed the opportunity to win big. It also meant it rarely lost big.

It was this same no-frills approach that Leininger encountered when he and his wife moved to Fort Worth. Even the executive offices in 1976 made no concession to the trappings of success. Located in a former factory just west of downtown Fort Worth, the headquarters of Lewis Kornfeld, then Tandy corporate vice president and president of the Radio Shack division, had a permanently waterlogged rug from a leaky ceiling. Add to this low-rolling attitude the fact that no one at the time was confident there was a market for a micro-computer, and the orders given to Leininger began to make sense.

The Profit Margin Of a New Machine

At first, the young engineer from California was told that he would have to design a computer that could be sold at under $200 retail. (The Altair, by comparison, sold for $397, and that was for a kit without a monitor and keyboard.) A $200 retail price translated to a design that cost no more than $80 to manufacture—$60 for parts and $20 for putting them all together. The original plan called for no monitor and a membrane keyboard.

After working on this plan for a while, Leininger went to Roach and French, who had become his boss, and pushed for added features that would raise the price. To be a workable computer, it would need a decent keyboard that would feel like the typewriter keyboards office workers were used to.

And it really should have a monitor. “That r
aised the retail price to $499,” Leininger says. “But that didn’t give Tandy the profit margin It was used to. Finally for $599 we were able to get the whole c
omputer with off-line storage In a cassette recorder, Basic built In with float- ing-point computation, with a screen and a good keyboard.”

The choice of the microprocessor was crucial. At the time, there were only a few likely candidates for the job: Zilog’s Z80, Motorola’s 68OO. Intel’s 8080, and a couple of chips made by National Semiconductors. Despite his familiarity with the microprocessors made by his former employer and a considerable knowledge he had acquired on the 8080, Leininger went for the Z80. It had what he considered a better set of instructions that meshed nicely with the Basic Tandy was picking up from Microsoft. (Microsoft’s Basic was chosen because it was already being used on the Altair, which at that time was enough to make it an industry standard in an industry that barely existed.) But the main factor in favor of the Z80 was its “automatic RAM refresh.” which would allow the new computer to use twice the memory that it could with other microprocessors—for the same amount of money.

This was a big achievement at this stage in the fledgling industry. Tandy’s computer would offer at least 4K of memory at the same time that 256 bytes were standard on the Altair and while other microcomputers rumored to be In development were only planning to have 2K. The Z80 also would allow the memory to be expanded even further with the use of 1 6K memory chips that would soon be commercially available.

The Blessing

Tandy at first was Spartan about staffing the project. For months, only four people, under Roach’s direction, worked on the new computer. French, who bought electronics parts for Tandy and was some- thing of an amateur computer hacker, was the on-site coach overseeing the day-to- day progress. Leininger was both designing the hardware and writing a make-do operating system (a more extensive operating system would soon be fanned out to an Independent programmer, Randy Cook). Van Chandler was pulled In from Tandy’s data-processing department to write applications software. And Dave Lien was writing a manual to accompany the computer’s Basic.

For months, the only ones who knew where the project stood were Roach and his team. Kornfeld knew that they were doing something with a computer, and he would sometimes stop by the converted used-car showroom where the new computer staff worked to see how things were going, Kornfeld would look at the tangle of wires and strange components that was supposed to be a computer.

“I’d say, ‘We’ll order it the day it can play chess with me and nobody from engineering has to hold the wires together or tell me “It was working Just a minute ago, but…” Kornfeld wrote in his book, To Catch a Mouse, Make a Noise Like a Cheese.

Around the beginning of 1977, after about six months of work, the team had a prototype of the computer ready to show off, and Roach called Kornfeld over to look at it. On his way to the windowless conference room where Roach’s team had set up the prototype, Kornfeld saw Charles Tandy by his black Continental. Until now Tandy didn’t know about the plans for a new product that Kornfeld and Roach were about to suggest they sell for a price higher than anything Radio Shack had ever sold before. Kornfeld figured now would be as good a time as any to let Tandy in on it.

With Tandy and Kornfeld giving this new contraption a hard look, Leininger

booted the computer, and it worked, without him holding any wires together or making any excuses. Of course, at this stage in the development, “working” didn’t mean that the apparatus did very much, certainly not play chess with Kornfeld. But Roach and his crew explained what the computer would be able to do- eventually— with the right programming.

” ‘Who wants a computer?’ Tandy asked

Kornfeld dutifully pointed out that they didn’t know if anyone would want to buy it. Except for some dedicated hobbyists, no one out there was demanding his own computer. What they had in mind was a computer for small businesses and schools, a market that didn’t exist be- cause the product had never existed. What’s more, Kornfeld added, the economics of buying parts meant that initially they would have to commit for 1 ,000 units.

Tandy was intrigued by the machine and said he figured that it would be worth the publicity even if the thing didn’t sell.

Roach then added that, actually, 3,000 would be closer to the number they would have to commit for.

Tandy thought this over for a while. Finally, they decided that since there were about 3,000 Radio Shack stores in the United States, if the new product was a bust, they could give one to each of the stores to keep their books on or to do inventory. . . or something.

The only person there who was convinced the project would be a winner was Leininger. He told Roach that even with the price at $599, he thought they would sell 50,000.

Roach’s assessment of that prediction was succinct.

“Horseshit,” he said. Tandy had never sold that many of anything at that price,

Leininger’s reasoning was that by the time you equipped an Altair with a monitor—the cheapest available was $795— you had spent more than $1,000. And the demand for Altairs, which were kits, was so great that they were months behind being delivered.

When Tandy left the room. Roach’s team still didn’t know if the project was on or not. It wasn’t until Feb. 2, 1977, that they got the official go-ahead. The new machine, which Kornfeld named the Radio Shack TRS-80 Microcomputer System, was scheduled to debut in New York on Aug. 3 the same year.

Saddling Up

Soon Leininger’s wife of less than a year learned to tolerate the 18-hour days he put in seven days a week. The team grew to seven people and moved into a former insulin factory where Tandy hi-fi speakers were made. For the final stage of the job, the actual building of the first units, the crew took over the upstairs floor of an abandoned saddle factory.

Some of the work was farmed out. At first none of the big video companies was interested in supplying monitors, possibly because Tandy, ever conservative, was asking them to supply only a few thousand. Finally RCA agreed to provide a 12- inch TV receiver stripped of its tuner, speaker, and assorted other circuitry. RCA also threw in a silver-gray cabinet that established the aesthetic design for the rest of the computer.

Tandy designed a custom keyboard, which also contained the guts of the computer, the microprocessor, and other circuitry. A separate interface unit that would allow the computer to be expanded with other circuit boards and external peripherals was in the works, but for the August debut, the staff concentrated on just the keyboard/processor, monitor, cassette recorder, and enough software so that a buyer could turn the thing on and actually do something with it.

Not long after the team began working in earnest, Roach made a trip that for the first time convinced him they might be on to something. In April, he went to California to attend the West Coast Computer Faire, a convention of computer hobbyists and what few manufacturers existed at the time.

“I saw 12-14,000 people, most of them payi

ng $9 a head and waiting in long lines to get in,” he says, “I thought maybe I really was looking at the rudiments of an industry.”

During the summer of 1977, the staff labored in the old saddle factory, hand-building the first 25 units in time for the kickoff to be held in just a few weeks at the Hotel Warwick. Then with only a month to go, Leininger ran into a problem he couldn’t figure out. The Invitations had already been sent to the trade publications, advertising brochures were already in the works, and Leininger was sitting in front of a TRS-80 that refused to communicate with the cassette recorder.

The 18-hour days were getting to Leininger. He was burned out. For two weeks he wrestled with the problem. Every time he made a change and hooked the computer to an oscilloscope, the machine tested fine. But when he hooked the computer to the cassette, nothing happened.

“I got to the point where I just didn’t know if I could figure out what the problem was,” Leininger recalls. “I told my boss about it. and he said he’d be in serious trouble if I didn’t get it to work.

“Finally after working on it all night, about 2am nature called. I went to the men’s room with a listing of the program. While I was sitting there, I found the problem.”

The New York Debut

When the TRS-80 was officially unveiled on Aug. 3, Roach was disappointed at the initial reaction by the press.

trs80

“Only a few radio hi-fi types showed up” he remembers. “The general-interest publications didn’t care about microcomputers at that time. The technology analysts clearly didn’t believe the microcomputer was anything of significance. They couldn’t relate to the tremendous gap be- tween standalone computers and minis.”

But while the trade press was underwhelmed at the debut of the TRS-80, the public was overwhelmingly enthusiastic about the computer. Radio Shack stores were soon flooded with orders—more than they could fill for months.

Harvard Pennington was typical of the type of person who saw what the TRS-80 could give him. Pennington would later make a lot of money writing books that explained how to get around some of the TRS-80’s flaws, and he would even sell a program called Electric Pencil, written by Michael Shrayer and one of the first word processors. Note that Pennington was not a hacker. The closest he had ever come to programming was working with a fancy Texas Instruments calculator.

Pennington used the calculator in an effort to cash in on the diamond market, which in 1977 had gone crazy. He had seen a program that ran on a mainframe that tracked the diamond market, but it was too expensive to rent the terminal time. He studied the program and figured out how it kept track of the many variables involved in trading diamonds. There was too much data involved for even a powerful calculator like the TI 57. That’s when he decided to get a computer.

Apples. Pets, and Choices

In 1977, Pennington didn’t have many choices for a personal computer (a term that hadn’t been invented yet). He didn’t want to build a computer from a kit, and only three ready-built computers were available: the Commodore Pet, one available only by mail order from Ohio Scientific, and the TRS-80. (The only Apple computer at that time was the Apple I, which was a kit: the fully assembled Apple II wouldn’t be ready for mass distribution until the next year.) Pennington doesn’t remember why he rejected the Pet. “Maybe it was the name,” he says.

He borrowed an Ohio Scientific for a few days, long enough to figure out Basic, but he didn’t buy one because the company refused to answer his questions when he called. (“Scientific dorks,” he calls them today.)

Then he went to a Radio Shack and saw the TRS-80.

“There was something about it”, he recalls, “A little bell went off in my head.”

Although the early TRS-80 was relatively primitive, Pennington was impressed by the add-ons Tandy was promising: more memory, a disk drive, and an Expansion Interface. A printer was also available.

“It had no tractor feed, just used rolls of paper without perforations, but it was a real printer.” he says.

Pennington placed his order, and after waiting several weeks, the computer was ready for him to pick up. Writing his first program wasn’t easy, but eventually, he got the job done and “it was a very good program,” he says.

The Deluge

Pennington wasn’t the only one who saw the possibilities of the TRS-80. At the time of the Aug. 3 debut, only 25 TRS-80s existed. Within weeks of the introduction. Radio Shack stores had taken thousands of orders for it.

“We were almost immediately deluged”, says Ed Juge, today the company’s director of market planning. He joined the company in the spring of 1978 while Tandy was trying to cope with the demand for TRS- 80s. “At one point we were nine months behind on delivering Level II ROMs, six months behind on disk drives.”

Both Komfeld and Roach took a lot of phone calls from irate customers demanding to know why their TRS-80s hadn’t been delivered.

“At the time we introduced them, we didn’t have the infrastructure to handle back orders,” Roach says. “We didn’t have the people for customer services. All that just evolved as the business grew.

“The only thing we could do was tell the customers everything known to man, and we were getting calls from legitimate customers, from people who wanted to use them in the state pen, and from people wanting to use them for things that they should have been put in the pen for”, he adds.

Roach would go down to the factory on Saturdays to help assemble computers. But no real assembly line yet existed. Each computer was crafted by hand, and output was only one a day. It wasn’t until March 1978 when the manufacturing staff had grown to 385 and the space taken up by computer operations had grown from 15.000 square feet to 85,000 that the company felt it had the situation under control.

Wives and the IRS

Today different people can find different, perfectly obvious reasons for the success that no one could have predicted before it happened. (Even Leinlnger’s guess that they would sell 50,000 TRS-80s was short by nearly 5,000 for the first year,)

“We caught the imagination of a lot of people,” says Juge. “[They] realized here was a way they could gain the same kind of management advantage in their little part of the world that others had had in big business for years.”

In 1979 Juge conducted a survey at the first of a series of barnstorming exhibits in 50 U.S. cities to show off the TRS-80, Seventy percent of the TRS-80 buyers claimed they were getting the computers for business.

“I guess th
at’s what they told their wives and the IRS,” Juge says now. “I think most of them just found it fascinating to use a computer. If they could figure it out and use it at work, then they wrote it off.”

Schroeder explains, “It was something you could slide by the wife. You could get an 8K machine for $599. Then buy the Level II Basic for 500 bucks. Then you told the wife that you needed one disk drive. Then maybe another. The entry-level price was low, but you could spend $4,000 over a period of two years.”

Roy Soltoff, who would later join Schroeder to write software for the TRS-80. saw other factors that continued to make the TRS-80 successful even after the Apple II became serious competition.

“The Apple all-in-one box didn’t appeal to the purist. And it had an inferior ROM for math. It couldn’t do floating point unless you got the Apple Basic language card,” he says.

As important as the hardware itself in the success of the computer was the ready-made distribution system through the thousands of Radio Shack outlets. They were more common than McDonald’s, and they were located in areas where the California-oriented Apple had yet to penetrate.

“Radio Shack had a better route to the people,” says Schroeder. “They had better dissemination of advertising material, better distribution, and a better repair and parts network.”

Trashing the TRS

No one claims the TRS-80 succeeded because it was perfect. In fact, even the most loyal TRS-80 owners called it “Trash -80,” a nickname earned by the computer’s inadequacies and frequent glitches.

Looking back. Roach considers the failure to plan ahead so Tandy could supply peripherals to be the biggest mistake the company made.

“It hurt us at a critical growth period,” he says.

To save the 97 cents that it would have cost to add some more memory to the TRS-80’s video display, Tandy left out lowercase characters. To get them, owners had to spend an extra $30 to buy an adapter kit.

There were other problems. The machine sometimes overheated, and matching the Expansion Interface to the CPU was often touch-and-go. But the biggest problems owners encountered were not In hardware but software.

Raiders of the Lost Disk

At one point Roach considered going with CP/M, the operating system created by Digital Research. It was becoming the de facto standard on S-100 computers, machines that traced their origins more directly to the Altair and used the same Z80 found in the TRS-80.

“But CP/M was an abortion.” says Juge.

“If you were a computer engineer, CP/M was just a set of input/output routines. It didn’t do anything to save you from your own mistakes. The manuals were virtually unreadable. I had programmed for three or four years, and I have more knowledge than most customers, and I couldn’t understand page 1.”

Still, the company bought the rights to CP/M as a backup in case something happened with the operating system it was developing in-house. And Roach warned a couple of times that if Tandy’s own product hadn’t reached a certain point by a certain date, he would go with CP/M.

Leinlnger’s original operating system was never designed to be more than a temporary package to get the computer off the drawing board. But the first full-fledged operating system, TRSDOS, wasn’t much better. Part of the reason behind the inadequacies of TRSDOS was the arrangement between Tandy and the person under contract to write TRSDOS, Randy Cook. Pennington describes Cook as a “bright but wary programmer who figured Radio Shack was out to screw him.”

Source: http://www.trs-80.org/
Source: http://www.trs-80.org/

At the time Cook was writing TRSDOS in 1977. Charles Tandy died, and the company was being run by committee, Pennington says. The company kept changing what it wanted from Cook, and the programmer, to protect himself, kept making changes in the operating system that only he knew about, Pennington says. The result was a TRS-80 user’s nightmare.

“It was a terrible time,” Pennington says of the year he used TRSDOS. “It was dreadful trying to do a backup with one drive. The backup and format software had terrible bugs in them.”

In the meanwhile others began writing alternative operating systems. Pennington saw a demo of one of these non-Tandy programs and was amazed.

“It didn’t crash!” he says.

But Pennington found that he couldn’t buy a copy of that program. APR-DOS, at his local Radio Shack. It wasn’t an authorized Radio Shack product. Eventually, however. Pennington got a bootleg copy of Newdos, another alternative operating system.

The operating system was handed over to Schroeder and Soltoff, who had written the unauthorized Newdos. The two Milwaukee computer enthusiasts had been quick to recognize the possibilities of the new computer and formed a software company called Logical Systems. Cook and Radio Shack bickered about the ownership of the TRSDOS code, and when It came time to write an operating system for the new Model II, the company took all copies of the source code away from the programmers. To avoid any further disputes with Cook, Radio Shack wanted to force the programmers to write a new operating system from scratch.

Today Juge admits there were flaws in the early attempts to convert Leininger’s original operating system designed for use with a cassette recorder to one for use with disk drives,

“It was admittedly a buggy version,” he says. “We told people this version is really not ready for consumption yet. But the people would say, ‘Tell us where the bugs are and we’ll try to work around them.’ “

A Hard Line on Software

Being a dreamer was almost a requirement for the early TRS-80 buyers. By the time the machine was introduced. Chandler had come up with only four programs for it. One program was for budget management. Another was a payroll program. The other two were hardly something you could use to convince the IRS you had bought the computer for business purposes: a blackjack game and a program to calculate your biorhythms.

“We got into some applications programs, but Roach didn’t want to be in the software business.” Juge says.

Roach says, “I didn’t think software was an issue. We weren’t terribly cognizant of what people wanted in software. We expected them to write their own.”

But at the same time Tandy shied away from getting into the software business, it refused to help others who did want to write programs for the TRS-80,

The Apple II was introduced with what was called “open architecture.” True to the thinking of the Homebrew Computer Club, Wozniak and his partner, S
teve Jobs, made no secret of how their computer worked. Third-party software companies sprang up and made fortunes writing programs for the Apple. The availability of so many programs, in turn, made the Apple II a more attractive machine. One of the programs written for it was Visicalc, the first electronic spreadsheet and the program that some have said changed the microcomputer from a luxury item to a business necessity.

Today Roach admits, “I wish Visicalc had been written for the TRS-80 instead of the Apple.”

But at the time, Tandy, halfway across the continent from the hacker ethic of Homebrew, took the more conventional big-company view that information about its products was proprietary. Radio Shack was the store that didn’t even sell flashlight batteries unless they had a house label on them,

“If someone at an Orange County user group wanted a copy of the source code for something that had, say, a Microsoft copyright on it, at the next usergroup meeting someone In a sweatshirt and blue jeans would pass out copies of it,” says Juge. “We had a little bit different way of looking at the world.”

Bennington contends: “Charles Tandy had wanted an open system for the TRS- 80, but when Tandy died, the committee took over. And it said, ‘If we keep control of it and don’t tell how it works, they’ll have to buy it all from us, and we’ll control the world.”

Although Radio Shack turned to Logical Systems to write the operating systems that became LDOS and TRSDOS 6, Soltoff subsequently was frustrated trying to sell TRS-80 programs he and Schroeder wrote independently of Tandy.

At first, Logical Systems was able to sell TRS-80 programs through Independent retailers. But that market dried up as the retailers found it too difficult to compete with the ubiquitous Radio Shacks and easier to specialize in software for the Apple and Commodore, Soltoff says.

Logical Systems turned to mail-order ads in magazines, but that was less than successful for the same reason: Radio Shack didn’t carry the magazines, severely limiting their circulation.

“Tandy has done a number of wrong things over the years, but if they had changed their position on this one point, things might be a lot different today,” Soltoff says.

Later, when IBM was to enter the microcomputer field, that same proprietary attitude and pride in the Radio Shack label was to help topple Tandy from the position it shared with Apple at the top of the new industry Tandy had been instrumental in bringing to full bloom.

Boom Times

The introduction of the IBM PC, though, was still four years away, and once Tandy caught up with the back orders for the TRS-80, Tandy went on a product binge, milking the newly discovered opportunity with upgrades to the TRS-80 and with entirely new computers.

tandy_trs80-model2_2
Model 2 and its 8″ floppy drives

In May 1979, Tandy introduced the Model II, which added larger-capacity, 8- inch disk drives and a faster Z80A microprocessor. It also corrected some of the more frequent complaints about the Model I (as it was now called): the difficulty of matching the CPU with the interface unit and its tendency to reboot when you least expected it.

hw-ad-model3-1The Model III, unveiled In July 1980, was inspired by new FCC regulations on radio-frequency emission, a standard the Model I flunked. With the Model III, Tandy made a radical departure from previous models. It combined the keyboard, CPU, Interface, monitor, and disk drives all in one cabinet. Except for a new reset button, inside the cabinet were the guts of a Model I, but the computer had become less of a hobbyist’s collection of individual parts and more a plug-and-run tool for the office and school. (The schools particularly liked the all-in-one design because it was harder for a student to walk off with than a disk drive or cassette recorder in his or her book satchel.)

Introduced at the same time as the Model III, the Color Computer was Tandy’s answer to the popularity of the Commodore Vic-20 as an inexpensive computer for home use. In March 1984, Tandy created the first successful laptop computer with the introduction of the Model 100. It was a battery-driven computer that fit inside a briefcase and that came complete with built-in word processing, BASIC, and the software and hardware necessary to hook the computer to a phone. The 100 became an instant necessity for reporters and executives who spent a lot of time on airplanes.

The Model 4 In April 1983 retained the Model Ill’s cabinet but upgraded it to a faster CPU and an 80-column by 24-line screen. It also included the TRSDOS 6 operating system, which finally answered most of the owners’ complaints.

By 1979 Tandy had sold more than 200,000 computer systems, topping $500 million in sales. The company was shipping hundreds more each day. There were more than 1,600 employees in six factories turning out TRS-80s alone. Computers and the seemingly never-ending list of accessories were quickly becoming Tandy’s biggest single source of revenue, growing from 8.5 percent of the company’s Income in 1979 to 34.5 percent in 1983.

During much of that time Tandy continued to run a nip-and-tuck race with Apple for the leading share of the market. In 1979, both companies had about 20 to 40 percent of the market, depending on whether you defined the market as home computers, business computers, or both. (One estimate for Tandy’s share at its peak was 60 percent.) But after that year, Tandy’s share of the market began to slip.

The decline in its market share was at first partially attributable to the fact that Tandy has never measured its success by market share. While other computer companies were cutting prices, Tandy refused to do anything that would cut into its profit margin. The strategy worked. Tan- dy’s profits, even while Its share of the market was eroding, continued to grow at the rate of 35 percent a year. But more and more the market wasn’t the same game in which Tandy was used to playing a dominant role. A new player had entered the game, and it had the clout to change the rules. The player was IBM,

The Big Blues

No one today will argue that the IBM PC introduced in 1981 was a radical advance in computer technology. IBM played it safe, using proven components that were in abundant supply. The most significant difference was that It used a 16-bit CPU, the Intel 8088, The 8088 meant that programmers could use up to 640K of memory for their programs. The 8-bit CPUs in Tandy computers and all other computers until the IBM machine came along were limited to 64K.

Just as importantly, the new 16-bit machines had the IBM logo on them. For a price not that significantly different from that for a TRS-80 or an Apple II, you could have on your d
esk or in your bedroom a computer made by the world’s biggest computer company. Executives, many of whom still considered the Apple something for kids to play games on and the TRS-80 something for hobbyists to tinker with, suddenly became a whole new market for computer sales. But the only one selling to them was IBM.

Today people at Tandy don’t like to admit it, but others are quick to say that Tandy was too proud to recognize the threat posed by IBM.

“The thing is that our market share declined precipitously,” Juge says of the first couple of years after IBM brought out its personal computer. “They caught us off guard. No one thought [IBM’s impact] would be as drastic as it was.”

By 1983, profit margin or no profit margin, Tandy’s share of the market was about half what it had been at Its peak In 1979. More importantly, its profits, which had been on a steady climb for four years, were fast leveling off.

Roach denies that there was any debate within Tandy as to whether they should jump on the IBM bandwagon, which was already loaded down with a hoard of established and new computer companies.

“Our long experience in the marketplace told us there were standards that evolved that need to be followed. In the 16- bit world, MS-DOS was the standard.” Roach says.

But it wasn’t until two years after the debut of the IBM PC, when Tandy finally brought out its first MS-DOS computer, that the Texans gave any indication they knew those standards existed.

“Our timing might have been affected by the overall size of the business we were generating with the 8-bit machines,” Roach says. “These machines continued to sell well despite the PC. There was no great urgency to create an MS-DOS machine.”

Roach may deny there was any Texan pride behind the slowness to make a concession to the standard being forced on the microcomputer game by this high- rolling newcomer from back East. But Tandy’s response to that standard, introduced In November 1983, still exhibited a stubborn streak. The Tandy 2000 was an MS-DOS, 16-bit computer all right, but it was compatible with the IBM PC only in the loosest definition of the word.

Source: 8bit-micro.com
Source: 8bit-micro.com

Not that the Tandy 2000 wasn’t a good computer. Even an experienced Tandy critic like Harvard Pennington was impressed by it.

“Technologically, engineering-wise, it was well-designed. It was faster. The way you took boards in and out—they slide on little trays—was slick. The Tandy 2000 even had a little bit of style in how it looked,” Pennington says.

The 2000’s Intel 80186 CPU was, in fact, two to four times as fast as the 8088 used in the IBM machine. And Tandy used drives that stored more information on a floppy disk than IBM’s drives. Graphics, which had been one of the most criticized features of the IBM were given more resolution on the 2000. Tandy even threw in a couple of extra function keys. Many who looked at the 2000 saw a better machine than the IBM PC. In a very real way, it was too good.

The Advantage of Mediocrity

When the IBM PC was introduced, computer programmers, being born hackers, quickly learned how to play tricks with the MS-DOS operating system, which they considered slow and cumbersome. The programmers figured out how to go around the operating system, which was designed to be a mediator between software and hardware so that their pro- grams could give instructions directly to the various hardware components. It was faster and, from the hacker’s standpoint, a more efficient way of doing things.

The problem with the Tandy 2000 was that if any of these programmer’s tricks tried to send instructions directly to the disk drives. CPU, or monitor, chances are the tricks wouldn’t work with the “improved” features. Tandy ran tests on its new MS-DOS computer with 100 of the most popular programs for the IBM PC. Half of them wouldn’t work. One of those was Lotus 1-2-3, which had justified purchase orders for IBM personal computers the same way Visicalc had legitimized the Apple II.

The programmers at Lotus ran tests on both the Tandy 2000 and the IBM PC. The 2000 finished in half the time the IBM took. The Lotus people told Juge “it was the neatest computer they had got their hands on.” They agreed to modify 1-2-3 for it, a job that took only four hours, Juge says.

But Lotus was just one software company, and an enormous advantage of having an IBM PC or a highly compatible clone was the wealth of software available under the IBM standard. There were too many software companies confronted by too many semi-compatible computers like the Tandy 2000 for all the programmers to keep up with what all the hardware engineers were doing. Like it or not, the minimum requirement for survival in the age of the IBM PC was to conform religiously to the IBM standards.

Other computer manufacturers— DEC, Texas Instruments, and Eagle—had already made the mistake of creating “Improved” PCs, but Juge says that Tandy hadn’t noticed that few people were buying the semi-clones.

“The reason we did what we did was that we didn’t like the idea of just going out and copying what someone else was doing. It was not our thing,” he says. “We knew we could either clone this computer or we could build the best damn MS-DOS machine our engineers could design— the state of the art in hardware.”

The spirit of jingoism was running so high at Tandy that Juge decided against suggesting they also build a clone “just in case.”

“I had nagging doubts [about the 2000], but not enough to argue with anybody.”

Lelntnger, who after his Model 1 bumout had left Tandy for a while and then come back, says he did tell his bosses the plans for the 2000 were wrong. But they didn’t listen to him.

“Tandy was under the assumption that there were going to be three standards for microcomputers: Apple, IBM, and Tandy,” he says. “I don’t know what they were thinking. The sun gets awfully hot out here. Maybe it baked their brains.”

The Standardization Game

Within six months of its introduction, no one had to tell Tandy the 2000 wasn’t going to fly, Juge says. “We knew we had made a mistake.”

A suspicion that they might have made a mistake may have begun to surface at Tandy even before then. The Tandy 2000 was introduced at the Consumer Electronics Show in Las Vegas on Nov. 30. 1983. Two months earlier John Roach had stood before a blackboard filled with specs for a new computer, one that was to be more closely compatible with the IBM PC.

“Gentlemen,” Roach told the engineers in the conference room as they studied the blackboard, “this is our next product. It is code-named August. I hope it’s obvious what that means.”

The plans were for the Tandy 1000, envisioned as a computer that would offer more functional compatibility with the IBM PC. but would still have hardware improvements. It would require fewer expansion boards by making a lot of the expansion board circuitry standard on the 1000’s motherboard. It would take up less space on a desktop, and despite the problems caused by the “improved” keyboard, speed, and graphics on the 2000 it was being designed with 12 function keys, would run faster than the IBM, and have a higher-resolution graphics display. This time, though, the graphics would conform to a different IBM standard, that of the PCjr.

While the 1000 was being readied for market, Tandy decided to take a second look at another computer design the company had rejected earlier. Jugi Tandon, head of one of the major manufacturers of disk drives for microcomputers, had earlier pitched Tandy on a 100 percent IBM clone his company planned to make. He wanted to know if Radio Shack would be interested in selling it under the Tandy label.

“At the time, it wasn’t what we were interested in doing.” says Juge, “but when Tendon came back to John [Roach] later on and said It was ready to go, we said, ‘Why not?’ People had been asking us to bid on large-quantity sales, and they were specing the PC/XT.”

Released in November 1984, the pure clone—the Tandy 1200—was not a success story. Juge says. The 1000, which was released a couple of months before the 1200, on the other hand, has been one of Tandy’s best-selling MS-DOS computers.

Tandy 1000HX - 8bit-micro.com
Tandy 1000HX – 8bit-micro.com

Since then Tandy has followed a policy of developing IBM compatibles that include a little something extra: extra speed, an extra expansion slot, an extra connection for a joystick or light pen.

Bouncing Back

The strategy seems to be paying off.

The installed base of Tandy computers of all types has grown from 1,045,000 in 1982— the year before its first MS-DOS machine—to 3,180,100 in 1986, according to International Data Corp., a market- research firm. Radio Shack stores sold 68.000 MS-DOS computers in the first quarter of this year, 62 percent more than the same quarter last year, IDC says. Another research firm, Infocorp. notes that last year Tandy sold 667.500 computers— not all of them MS-DOS machines—putting it in a tie once again with Apple, each claiming 25 percent of the microcomputer market. The difference now is that the two companies are tied for second behind IBM.

What’s more important to Tandy in the long run is that it no longer pretends it can ignore the presence of IBM. In 1986 it began emulating another well-known IBM standard: the well-dressed salesperson. After inspecting several Radio Shack stores and coming away in a state of shock from some of the grungy furnishings and grungier sales staff, Roach issued a directive establishing the company’s first dress code. Roach also began a store-by-store re-furbishing designed to vanquish once and for all the Radio Shack image as a hobbyist’s haven. He replaced it with a new image as “the technology store,” specifically technology catering to businesses. He also began a program of more training for computer sales personnel and established an outside sales force to reach the business executives.

And Tandy is no longer shy about IBM technology. Although he refuses to be specific. Roach says that it will match IBM’s move earlier this year in creating a computer based on the powerful Intel 80386 CPU.

(The people at Tandy take some satisfaction from the fact that IBM’s new line of Personal System/2 computers has a 12- function-key keyboard and a new type of expansion slot that resemble those on the hapless Tandy 2000 more than they re- semble those on the IBM PC.)

Roach welcomes the changes in the IBM products because their new disk drives, analog video, and “microchannel” bus introduce hardware incompatibilities with IBM’s own older PCs. The changes free Tandy from what rankled the company in the first place, a slavish adherence to IBM hardware.

“OS/2 [the new operating system under development by Microsoft for the 80386 computers] is hardware independent,” Roach says. “Windows is hardware independent. What you’re really saying is that from a hardware standpoint, the manufacturers are going to have the opportunity to provide different features, different performance—if they will run the same software,”

Roach also hints that Tandy has an MS- DOS laptop in the works, another area in which Tandy has lost its lead to MS-DOS computers. In this case the high-power portables from Japan.

On the other hand, the Model 4, at one time Radio Shack’s biggest-selling computer, is doomed.

“We’ll sell the TRS 8-bit machine as long as there’s a continuing demand for it,” Roach says. But he adds: “Certainly its popularity and volume is continually declining.”

He has better hopes for the continued success of the Color Computer line as a home computer, although he refuses to call it a home computer.

“One of these days, the industry will develop a true home computer. It’s something our people are working on.” he says without elaborating.

Overall, there is a sense of direction, a sense of organization at Tandy now that didn’t exist 10 years ago when Roach noticed his engineers tinkering with the first microcomputers. The way Roach talks. Tandy won’t be caught again either napping or sticking its head in the sand:

“Our goal now is selling all technology products that have a broad customer base—and in many cases being on the leading edge of those products, such as we’re doing developing the cellular phone market, which is really not a market yet. That’s not something we’ve always done with microcomputers. But that’s not a mistake we’ll make again. “

This article was first published on 80 Micro magazine, issue August 1987 and reproduced here with the “blessing” of its author, Ron White. The article was revised and illustrated to offer an interesting perspective on how Tandy went from that point in time until today. The story was chosen as the best feature among all the CPM publications that year.

Ron White is the former executive editor of PC/Computing magazine and the author of the best-selling How Computers Work 10th Edition The Evolution of Technology. The book’s Facebook page is at http://tinyurl.com/HCW10th

Author: Paulo Garcia

4 thoughts on “The Story of Tandy – 10 years after the TRS-80 Model 1

  1. Sorry. got lots wrong. Twin Granite Towers? We did not move into them until after the TRS-80 was released. The bottom did not drop out of the CB market until the same time the TRS-80 was released. It is really what saved the company. CB was a large percent of their business. They were NEVER looking for something. When I first started attempting to get their interest I was laughed at. Even Jon Shirley (Former president MicroSoft) who was the Merchandising VP for Radio Shack International told me “Don’t waste my time, we can’t sell computers”.

    John Roach as all executives were against the idea. Bernie Appel (VP Merchandising and the executive directly over me) at one point had is secretary tell me “If you are working on that GD computer, you are fired?”

    The 1,000 number was correct number. The 3,000 number did not come until many months later. When Mr Kornfeld said we were upping the number the production to 3,500 so that WHEN the product fails, we will use it in the store for inventory control. Steve and I both thought he number would be 50,000 first years and we were laughed at.

    The West coast computer faire was before they made the decision about the production number. It was what gave Roach the realization that there really was a market.

    I was initially working toward the “Project Board” approach since I had control over it. I could add it without getting any approval from any executive. As a Product Line manager I could put what I wanted in the line since I was responsible for the parts and pieces the company sold. If I felt it was right, I could add it. I did want the true personal computer and knew what it was. See below about the Pet.

    We did not go with CP/M when we added the disk system because we used the low interrupt address in the ROM. It would not work. I introduced the first CP/M for the Model I after I left Radio Shack by changing the addresses and worked with Bill Gates to port all of his software to it. (I had hired Bill to write the Level II basic at Radio Shack. That is another story.)

    An aside about the Pet. Chuck Peddle was a vendor of mine at Radio Shack. He was in my office one day before Radio Shack even accepted the idea of a computer and I was under no non-disclosure as it was my ideas and work even before Steve came to Fort Worth. I was explaining to him my concept of a personal computer. It was the Pet. Even down to the Microwave keyboard that the original Pet used.

    A lot of people believe that Apple created the Personal Computer Revolution and that is wrong. Apple had no money and no consumer computer and would have been bankrupt if it had not been for the TRS-80. They had no way to create the market. It took the marketing power of Radio Shack to do that.

    THE TRS-80 STARTED THE PERSONAL COMPUTER REVOLUTION.

    There was no other company who had the financial ability nor the market clout to do it at that time. If Radio Shack did not do it then, who knows when it would have happened.

    1. Hi Donald

      thanks for the long comment, clarifying some of the points of the article. It is interesting to see the point of view of an “insider”, which is different than what any journalist could get. I really appreciate it!

Leave a Reply

Your email address will not be published. Required fields are marked *