|An Introduction to the BPA
By Bob Fabris
Bob Fabris was the publisher of the "Arcadian" newsletter from 1978 to 1986. In the absence of Bally Professional Arcade coverage in most magazines, the "Arcadian" was one of the few ways in which most users were able to communicate with each other. In early 2001 I asked Bob to write an introduction for this Astrocade CD, this is what he has to say.
In the mid-'70s, the Bally Manufacturing Co. went modern. The company's history had started with a wooden 1931 pinball machine; at that time, it had been the leader in both pinball and slot machine production, claiming over 70% of the market.
But now the electronic age was coming, and in 1976, Bally created a computer-controlled pinball machine for home use: "FIREBALL." As an adjunct, the company also went into the home computer business, developing a machine in early 1977.
The computer know-how came from Dave Nutting Associates (DNA), a company that had marketed "COMPUTER SPACE" in 1971. This was the commercial version of the first video game ever, "SPACE WAR," which had been developed in 1962 by a team at MIT using Digital Equipment Corporation's PDP-1 mainframe. SPACE WAR had been freely distributed amongst other university groups for years. Jay Fenton of DNA was responsible for "FIREBALL," even managing its final design. Manufacturing was done by another subsidiary, the E.F. Johnson Co. of Garner, IA. Bally set up a Consumer Products Division to oversee the home entertainment operation. The 1977 home computer was to be a two-part unit. The first was a game console attached to a TV set, while the second, to be introduced shortly afterward, featured additional memory, a keyboard and full computer features.
The console contained a Z80 microprocessor by Zilog and an alphanumeric keypad (calculator-sized). Two hand controllers were included, along with connections for two more. An entry slot accepted cassette-sized cartridges containing one or more games apiece; storage space for these cartridges was also provided within the console itself. With 8k of ROM and 4k of RAM, the unit could be used immediately, as there were two built-in games, a calculator and an Etch-a-Sketch-type drawing utility. Output was NTSC on channel 3 or 4; a cord connected the system to the antenna input on a standard television. 256 colors were available, as well as a music synthesizer with six voices. Game cartridges, purchased separately, contained PC boards and connectors, with ROM chips of up to 8k. Games programmed onto these chips were usually based on Bally's full-sized arcade games.
The second unit, known as the "Add-On", was to contain a full-size typewriter keyboard with additional memory on the order of 20-32k, and a new language to allow the user to enter programs himself. Delivery of these units was to follow the base console by about two weeks.
Preproduction units were shown to Joseph Sugarman, of the Chicago high-tech mail-order house JS&A. This company was an early version of the stores epitomized by Brookstone and Sharper Image. After two months of writing and rewriting, Sugarman created a two-page advertisement extolling the machine's virtues. It was placed in the Wall Street Journal, Scientific American, Popular Science and airline magazines in September and October of 1977. A reading of the ad today is still as intriguing and compelling as it was then. JS&A promised a modem, a printer, the Add-On and a self-test cartridge. The modem (Dial-A-Bargain) was to be a freebie for customers who had bought consoles prior to March 1, 1978, allowing users to order game cartridges directly from JS&A; but it never showed up. Nor did the printer. The Bal-Check cartridge didn't appear until July of '81. The Add-On was to have a dual cassette deck for program storage and retrieval, along with the full keyboard. Specifications later changed to provide cassette jacks instead of decks.
JS&A was the exclusive outlet for the "Bally Professional Arcade" (BPA), or "Home Library Computer," and 8,000 orders soon piled up. At that time, the BPA was to be delivered in four weeks, but it actually started in January of 1978 (the expansion unit never showed up at all). The base unit was plagued by overheating, which resulted in many units being returned to the factory for warranty repairs. On a visit to the factory, I saw a pile of motherboards that was eight feet high.
On receipt of my unit, I could see that the new social concept of a User Group, where people could learn from each other, would apply nicely to the BPA. To start with, I placed an ad in ON-LINE, a computer-oriented "sell'n'swap" paper, in March. I then wrote a sort of round-robin newsletter for five months, with the content being almost entirely from contributors; readers provided news and rumors, and some experimenters opened the box and reported their discoveries (the 50-pin expansion connector, the light pen socket and other goodies).
After months of continuous delays with regard to the keyboard/expansion unit (Add-On), a stop-gap was produced in the form of a "Tiny BASIC" cartridge (6002), providing users with the ability to create their own programs by allowing access to the Z80 CPU. The language was a sub-set of Dr. Li-Chen Wang's "Palo Alto Tiny BASIC." Two contemporary documents contain material on this language: "Dr. Dobb's Journal of Computer Calisthenics and Orthodontia" (first-year volume) and PCC's "Reference Book of Personal and Home Computing."
The base unit was now termed "Level One." With the addition of the Tiny BASIC cartridge, it became "Level Two".
The Tiny BASIC cartridge was shown at the Summer Consumer Electronics Show in June of 1978, but delivery was not started until October. This became an appropriate time for my change to a subscription format; I then placed another ad, this time in Wayne Greene's KILOBAUD magazine. The first subscription issue of the ARCADIAN appeared on 11/6/78 and contained the first words about the internal operations of the BASIC language, courtesy of an included article by NCE/Compumart. Also available with the cartridge was a cable set that would plug into both the light pen outlet and a hand controller socket on the BPA at one end, and into a regular cassette recorder at the other. Programs entered into the machine could then be saved, stored and retrieved. The cassette could of course be used on another machine.
The ARCADIAN began running tutorial articles in Issue #2, and subscribers began to submit their own programming efforts for publication. In order to stimulate this activity, I published practically everything that was submitted. By studying the program listings and experimenting, one could learn how various effects could be created.
Work continued on the Add-On by the programmers at DNA. The language for this unit, while originally intended to be BASIC, was changed to GRAFIX, which accentuated graphics capabilities. Developed by Dr. Tom DiFanti and Nola Donato of the University of Illinois, along with Jay Fenton of DNA, it was a rewrite of an earlier language, GRASS32, which had been written by Dr. DiFanti for his 1974 PhD. thesis at the University. GRASS had been utilized by Larry Cuba, amongst others, for some of the special effects in the movie STAR WARS. GRASS worked on a PDP computer and had to be revised for use with the Z80.
By mid-October of 1978, JS&A was very unhappy with Bally - the heating problem, a delay by the FCC in providing approval for the Add-On, lack of the Add-On itself, and finally the non-exclusivity of sales. A change in Bally's management resulted in a dealer network being set up, contrary to the JS&A agreement. A third of the product went to the Montgomery Ward chain. JS&A sent two letters to their customers in which alternative courses of action were discussed. In one, JS&A proposed taking over the production of the Add-On if Bally did not produce it; in the other, JS&A revealed that they were negotiating with an entirely different company for the sale of a similar two-part computer/game system. We never found out who this new player was; Bally kept promising the delivery of the Add-On, so JS&A never became involved with its production.
Bally showed a version of the Add-On, now termed the "Level Three System," at the January '79 CES. It was boasted as including 20k of RAM, 32k of ROM, and the GRAFIX language. JS&A got fed up with everything in November and commenced with a "fire sale." Arcade units were let go for $49.95; 3,256 cartridges were unloaded at half-price. They also tried to sell 60 defective units at two cents each, saying that they could be repaired for $25.00 apiece under Bally's warranty policy. This wound up in court.
My full-time job as a systems engineer with Lockheed required frequent cross-country trips. I was able to influence meeting dates, which enabled me to attend conventions and meetings that featured Bally participation -- usually on weekends. As a result, I had many opportunities to meet Bally and DNA personnel at various CESs in Chicago and Las Vegas, and this helped in the production of the ARCADIAN. I was also able to have a booth at various West Coast Computer Fairs in San Francisco. By renting a couple of TVs and bringing the BPA, local subscribers and I could show off various games, etc.
In producing the ARCADIAN, I was very ably assisted by a number of local computer whizzes (i.e. Tom Wood, Dick Strauss, Dick Hauser and Al Rathmell, amongst many others) who frequently set me in the right direction. Production of the ARCADIAN was, however, a one-man job, with the organization and pasting up of the material done over each weekend. The 11x14 sheets were given to the printer each Monday to reduce, lithograph, collate, staple and fold the pages, and I picked up the finished products on Thursday, ready to address and group them according to USPS requirements. They were handed over to the post office each Friday. One subscriber paid for his subscription by supplying me with address labels (modified every month). A hand-wired Apple II was used for the text, and a Comprint electrostatic printer provided the output (both from the BPA and the Apple). A few mistakes crept into the early issues, since all programs initially had to be hand-listed. Typos and such were pretty well eliminated when we were able to print listings directly from the BPA. In the middle of Vol. 2, a program was successfully transmitted by telephone between two tape recorders and then inputted into the BPA, providing the most direct transfer mechanism.
It was discovered early on that the BPA could be connected to a video monitor instead of a TV, and a schematic was provided in the ARCADIAN for the small change required in the console's hardware. Unfortunately, the output was only black and white; but about two years later, a technique was presented that kept the color.
Toward the end of 1979, a number of Bally distributors and retailers began to desert the team. Corporate Bally was looking in a new direction -- casino operations -- and had bought some real estate in Atlantic City. Accordingly, the Consumer Products Division became "available." After a couple of false starts, the whole operation was taken over by a start-up company called Astrovison in August of 1980. Again, the Add-On was promised, this time including the ZGRASS language, which had internally supplanted the original GRAFIX. While GRASS/GRAFIX could provide 2D object animation, ZGRASS was able to provide 3D. The BPA was renamed the Astrocade. Dan Dawson, an Astrovision vice-president, spearheaded most of the public relations activities, but after meeting the team, I soon saw that the Sales VP, Ray George, was the real power behind the throne and called the shots. George had convinced an investment group to purchase the exclusive rights to manufacture and sell Bally products for a five-year period, for a cost of $1,650,000 plus royalties.
One of the first new Astrovision products was Jay Fenton's rewrite of the BASIC cartridge. 6004 now incorporated the tape input/output facility in the cartridge itself. In addition, the speed of transfer was raised from 300 to 2000 baud, and an LED was provided to show transmission activity. The program "Artillery Duel," which had appeared in the ARCADIAN in 1979, was chosen as a sample program in the new BASIC Manual. Unfortunately, the Manual's listing of the program had an error (a correction was printed in the ARCADIAN). One of Artillery Duel's writers, John Perkins, became an advisor to DNA, and in that capacity he worked on revising the program to commercial cartridge form (5005).
An extensive discussion of the GRASS language by the authors had been provided in Issue 2 of the ARCADIAN. This language was useful for commercial artists, movie animators, classroom instructors, flight simulation designers and others. On the commercial side, the computer design itself evolved and was eventually marketed as the Datamax UV-1, with price tags in the $2,500-$5,000 range. It was shown at various SIGGRAPH conventions over the years. In the popular press, another description was published in the June 1981 issue of Creative Computing, while a user provided his comments in the November issue.
While the included RAM had a "4k" capacity, the actual memory that could be written to was 1.8k. Many space-saving techniques had to be used if worthwhile programs were to be created. But in 1980, various subscribers began to consider their own expansion ideas, and by mid-year we saw the beginnings of memory addition schemes. These were supported by articles in the ARCADIAN. In addition, through the efforts of Brett Bilbrey, George Moses and Bob Weber, we made a breakthrough in learning how to utilize the three-tone music capabilities.
As subscriber-submitted programs became more sophisticated, we produced a "Sampler Tape" of ten programs of various types, later providing a "Best of Arcadian: 1980" tape. Authors received royalties based on the sales of these tapes.
Also in 1980, competition with the ARCADIAN appeared in the form of CURSOR (aka BASIC EXPRESS), a similar newsletter, based in Los Angeles. After printing 15 issues (and proposing a repair service for the BPA), the publisher moved to an unknown address in the Sierra foothills.
At a local Astrovision meeting in May of 1981, one of my valued subscriber-assistants, Dick Hauser, suggested to Dawson that Astrovision sponsor a prize for the best subscriber-invented program appearing in the ARCADIAN. After a couple of months of publicity, Astrovision backed out; the $100 prize monies had to be provided by the ARCADIAN. The general arrangement provided for at least three contestants to be judged by a panel of five subscriber-experts. The winner would then replace the most senior judge for the next contest. Dick went on to publish the SOURCEBOOK, an annually updated index to all programs, from all sources, ever produced for the system.
Another discovery was that there were two different ROM configurations in the field. The reason was never disclosed, but a short program was published that allowed the user to identify which chip he had. The difference meant non-compatibility regarding program transfer between ROM types.
Virtually all cartridges were invented/developed by the programming team at DNA, who also provided Bally with the programs for their full-size game machines. A very talented group, they were anxious to get the BASIC and later GRAFIX/GRASS languages out. Under the leadership of Jay Fenton, they designed and implemented the operating system for the BPA, later redesigning the BASIC cartridge and creating the Extended BASIC used with the Viper and Blue RAM memory additions. Jay continued to work on "Hot Rod Bally BASIC" and a "Color BASIC," neither of which ultimately went into production. Other people on the DNA staff were Jeff Frederiksen (the Projects Manager), Bob Ogden, Rickey Spiece, Larry Cuba, Scot Norris and Dick Ainsworth. Also involved through the U of I were Frank Dietrich, Jane Veeder and Zuszsanna Molnar.
Meanwhile, some ARCADIAN subscribers were interested in producing their own material for sale to Astrocade owners. A number of third-party vendors sprang up and produced games, utilities and memory additions: Esoterica, Bit Fiddlers, L&M, New Image, WaveMakers, Spectre Systems, Tiny Arcade, Craig Anderson, Alternative Engineering, Perkins Engineering, George Moses, Dave Ibach, Steve Walters, Bob Weber, Mike White, Bob Wiseman, SuperSoftware, Sebree's, R&L Enterprises, and Dave Carson, among many others. The Bit Fiddlers produced a cartridge that allowed programmers to bypass BASIC entirely and input machine code directly into the Z80 CPU. The ARCADIAN began to pay for tutorial articles.
In November of 1981, Astrovision moved their corporate offices to Rancho Cordoba, California (near Sacramento) and set up a production facility at Sierratronics, Inc., in order to augment the Garner plant. This was a surprising move to those of us outside the Astrocade family, and came about through the influence of James Guerin. Guerin had been one of the original Astrovison investors (Sierratronics was one of his interests), and he was CEO of Astrovision around this time.
1982 was a bad year for Astrovision; sales were up $2 million over 1981, but about $4 million was tied up in disputes. In May, the Quaker Oats Co. purchased a 13% share in Astrovision for $3 million, which certainly helped the cash flow; but the company later reneged, utilizing their option to back out (Astrovision had to eat the $3 million). The company was going to change the name on the console from "Bally" to "Astrovision" as soon as its right to Bally trademarks expired, but name conflicts with other companies arose, at which time the company decided on "Astrocade" -- after already spending yet another $3 million on advertising and promotional material, which all had to be essentially redone.
Early in 1982, Nitron of Cupertino, California became another player in the Astrocade group. In June, it became the principal supplier of hardware to Astrocade. In a rather unusual arrangement, Nitron bought Astrocade's entire arcade, cartridge and component inventory, contracting to purchase component parts (the major chips were sole-sourced from American Microsystems). Nitron then assembled the gear and sold it back to Astrocade. The purchase gave Astrocade some needed cash but made it highly dependent on Nitron's performance. Some Bay Area subscribers and I visited the facility and were impressed to see that Nitron apparently had all of the necessary equipment and capabilities to complete all of the required work under one roof (although the Garner and Sacramento plants were to continue their efforts, with Nitron as overseer).
In mid-'82, Perkins Engineering, which had been providing a memory addition called the Blue RAM, was contracted by Astrocade to finalize the Add-Under. This involved the creation of a printed circuit motherboard (all previous versions had been hand-wired and soldered, or had used wire-wrap techniques -- hardly an efficient method of production). In separate negotiations, Alternative Engineering obtained the rights to Dr. DiFanti's follow-on language ZGRASS for use with their keyboard addition unit the Viper System, which appeared in December of 1983. The Viper was priced at $495 ($1,495 with the language and a disk drive).
In October, Nitron took control of Astrocade via a stock trade; in December, Astrocade went into Chapter 11 Bankruptcy. Third-party vendors began to consider the use of cartridges for their output instead of tapes, and a number of programs were produced in this format. The DNA team had written an improved version of "Pac-Man" called "Muncher," but legalities prevented its retail appearance. Cartridges were available only to ARCADIAN subscribers.
In mid-'83, "Treasure Cove" by Esoterica appeared. It was the first vendor cartridge to incorporate a "Tournament Level," which allowed two or more players to compete separately within the same set of game parameters.
Astrocade emerged from bankruptcy in late '83 with a reorganization plan. One area under consideration was foreign sales; R&D work had been underway to develop chips that would operate with 50-cycle, 220-volt power supplies. These chips would be compatible with both PAL and NTSC video formats. Arcade units and cartridges were to be produced in Korea by Daejin Audio. But the plan resulted in Astrocade's folding under Chapter 7 in 1984, and it was Nitron's turn soon thereafter. After a close-out sale of $5 million worth of production equipment in March of 1985, they too filed, effectively closing the door on the entire project. Nitron arranged with a local company to sell off the available cartridges, and the ARCADIAN provided a mailing list of possible purchasers. A year later, Alternative Engineering stopped production.
After six years of publishing monthly issues, the ARCADIAN went quarterly in 1985. Around this time, Mike White developed a scheme by which taped programs could be easily transcribed into cartridge format -- but it was too late. The next four issues of the ARCADIAN were spread out over two years, and we closed shop in August of 1986 with a discussion of the Music Maker cartridge that had recently appeared. The subscriber base was continually diminishing, and there were no new BPA sales to generate new owners (we had a flyer in the box). Of the forty-two cartridges publicized by Astrocade, ten apparently never made it, while another seven were presented after Astrocade's demise. Its financial difficulties came about as a result of Nitron not providing product, which limited sales - as well as strong competition from Atari, Coleco and Mattel. In addition, there were five lawsuits concerning monies that had not been paid to Astrocade for product, as well as nineteen suits against Astrocade for various reasons -- all which obviously drained finances. With hindsight, what would the situation have been had Bally produced the Add-On in a timely manner?
Production of the ARCADIAN was very satisfying. It increased my knowledge of the computer world, made it possible for me to meet some very interesting people, upgraded my business skills, allowed me to play with new toys, and provided me with a warm feeling in knowing that my efforts had enabled some subscribers to become entrepreneurs in designing and selling programs and hardware (and even making career changes, in some cases). My hopes are that this was, for them, just a stepping-stone toward better things.
Robert Fabris - April 13, 2001
Paul Thacker found a draft of this article among the Bob Fabris Collection. If you're interested in checking it out too, then you can read it here: