Stan Sieler described the HP Amigo for Wikipedia:
The HP 300 “Amigo” was a computer produced by HP in the late 1970s based loosely on the stack-based HP 3000, but with virtual memory for both code and data. Designed as a single-user workstation, it was a commercial failure, massively so, considering the huge engineering effort.... The circuit boards were in a floor pedestal, with CRT and fixed keyboard on top. It pioneered such ideas as built-in networking, automatic spelling correction, multiple windows (on a character-based screen), and labels adjacent to vertically stacked user function keys, now used on ATMs and gas pumps. It also featured HP-IB as the I/O bus, an 8” floppy disk, and a fixed 12M Winchester hard drive. [i]
Two issues of the HP Journal – June and July 1979 – had fifteen articles penned by twenty-two authors that described the new system.[ii] Highlights were described as a “personal computer designed for the individual worker”, with guided “soft keys” for easy user-interaction, “on-board everything” including display, built-in hard and soft disc memory, and a stylish package. Sixteen machines could be easily networked. Priced slightly higher than the Alto at $38,500 with matching specs and a different, much easier interface by conventional wisdom of the day, it was a major statement by a major competitor already in the computer business. As noted multiple places, “the competitive landscape of the era was dominated by costly mainframes and minicomputers equipped with time-shared dumb terminals. Amigo was launched three years before the Xerox STAR; personal computers (circa 1981) were simplistic, with limited processing power and the inability to communicate with other systems.”[iii]
Aside from size, price, portability, software, and success, it was everything a personal computer should be. $40,000 for a personal computer might seem laughable, but the HP 9845C from Fort Collins two years later, a color-display workstation sans networking, sold like hotcakes at $39,500. Ironically, the $33,000 HP 250, characterized as a small-business computer, was developed simultaneously by the Böblingen Desktop Calculator group in one-fifth the time for one tenth the R&D investment, and it outsold Amigo tenfold. But it wasn’t politically correct – re-assigned to Cupertino under Bill Krause, it died of inattention. The HP 250 Journal article appeared two months in front of Amigo, two years after debut. Such internecine warfare was unbecoming of a well-managed corporation, but it increasingly characterized HP Computing. Something had gone radically wrong.
[i] Wikipedia HP 300 http://en.wikipedia.org/wiki/HP_300; http://www.answers.com/topic/hp-300
[ii] HP Journal (30:6); Jun 1979; Cover: HP 300 Computer; George R. Clark, “A Business Computer for the 1980s...a totally new business-oriented design based on HP's silicon-on-sapphire integrated circuit technology, this new system packs a vast amount of processing power into a surprisingly small package,” pp. 3-6; James R. Groff, Eric P. L. Ha, “The Integrated Display System and Terminal Access Method...the HP 300 handles up to 16 application terminals simultaneously. Its own display can act like several mini-displays at once,” pp. 6-9; Frederick W. Clegg, “Reducing the Cost of Program Development...it's a compiler-based system, so run-time efficiency is high, but it has many of the conveniences of an interpreter-based system,” pp. 9-15; James R. Groff, Phillip N. Taylor, Alan T. Pare, “Managing Data: HP 300 Files and Data Bases ...choose one of seven different file structures or the IMAGE data base management system,” pp. 16-19; Tu-Ting Cheng, Wendy Peikes, “An Easy-to-Use Report Generation Language...templates on the screen take the place of RPG coding sheets,” pp. 20-23; May Y. Kavalick, “HP 300 Business BASIC...it's specially designed as a versatile business applications language,” pp. 23-26; David A. Horine, “Innovative Package Design Enhances HP 300 Effectiveness...monocoque construction is the starting point and even the shipping container is novel,” pp. 26-30; Ronald E. Morgan, “World-Wide Regulatory Compliance,” p. 30. See also HP Journal (30:7), July, 1979; Cover: HP 300 Computer; Arndt B. Bergh, Kenyon C. Y. Mei, “Cost-Effective Hardware for a Compact Integrated Business Computer...CMOS/SOS technology helps reduce an eight-board processor to only two boards,” pp. 3-8; W. Gordon Matheson, “A Computer Input/Output System Based on the HP Interface Bus...it's designed to make it easy to add, delete and communicate with peripheral devices,” pp. 9-13; Richard L. Smith, “A Small, Low-Cost 12-Megabyte Fixed Disc Drive...a new Winchester-type disc was designed to meet the mass memory needs of the HP 300,” p. 11; Alfred F. Knoll, Norman D. Marschke, “An Innovative Programming and Operating Console...windows and softkeys add new facets to the classical concept of interactive programming,” pp. 13-17; Ralph L. Carpenter, “AMIGO/300: A Friendly Operating System...an improved man/machine interface sometimes called friendliness, requires an advanced operating system,” pp. 17-24; James C. McCullough, Donald M. Wise, “Configuring and Launching the AMIGO/300 System...system generation and startup are easier than they used to be,” pp. 20-21; Thane Kriegel, Dilip A. Amin, “A Multiple-Output Switching Power Supply for Computer Applications...designed for computer mainframes, this OEM power supply is an economical solution for the HP 300's power requirements,” pp. 25-28.
[iii] Xerox Star, Wikipedia, http://en.wikipedia.org/wiki/Xerox_Star
9 comments:
By 1980, what had gone wrong with HP computing, from my perspective in Colorado, is that HP corporate let the various computing division managers run their divisions like kingdoms. Once the Colorado Calculator Products Division established the market for desktop computing, every possible division jumped in. Corvallis introduced the HP 85, growing from handheld calculators to desktops. The Data Terminals Division added another processor and came out with a terminal-based desktop computer. There was Amigo and SAM, the HP 250, as you've discussed. Plus the growing number of models from the newly renamed Desktop Computer Division in Colorado. All of these machines were mutually incompatible in tape format, language, and networking abilities. It was a mess. Something had indeed gone wrong and in my opinion, it was a lack of direction for a corporate computing identity.
Chuck House wrote:
"Such internecine warfare was unbecoming of a well-managed corporation, but it increasingly characterized HP Computing. Something had gone radically wrong." Steve Leibson's reply agrees. I disagree (for the most part :-).
At the time, computers were rapidly moving from their raised-floor computer room isolation to businesses & homes. No one understood the markets, and HP's multiple investments in similar products (HP 125, the terminal which ran BASIC vs. desktop computers vs. the HP 250/300) allowed the market to figure out what it wanted & to vote with its cash. I knew the HP 250 design engineers in Fort Collins & they were all enthusiastic about their product, holding weekly meetings & giving "Conehead awards" to those who made schedule-delaying mistakes. The HP 250 software
& hardware was leveraged from
(=mostly based on) existing
desktop computers, mostly the 9830
& 9831, whereas I heard that the
HP 300 software was mostly new
& still had user interface issues.
In my opinion, the only thing that went wrong was that Paul Ely 'punished' the HP 250 team for their marketing success by transferring their product to Cupertino (and, from there, quickly elsewhere).
It should be noted that the marketing wars in HP weren't as bad as the gridlock in IBM in those days. IBM's solution was
to take a few good engineers,
send them to Boca Raton,FL,
tell them to develop a competitor
to the Apple II, TRS-80 & Commodore PET, then leave them alone. Specifically, IBM division managers were great at politics, but independent Boca Raton division succeeded because -- and only because -- the new IBM PC didn't have to be compatable with any of the previous computer lines.
chuck said...
volumes could be written (and have been, in fact) on the various perspectives from IBM, DEC, and HP on these fronts. The post here was primarily intended to illumine the HP 300 only, rather than the host of machines HP built. The question with the HP300 vs HP250 did turn out to be internecine warfare, which will be treated in the book. But the Amigo itself was really a precursor for the Xerox Alto, even the MacIntosh. And it shared some of the same developers with the Mac, notably John Couch and Jim Groff.... So, understanding what went wrong with Amigo might help understand how HP missed the PC shift???
Ah, missing the PC shift has little to do with Amigo in my opinion. Desktop Computer Division was defining itself as putting machines on engineers desks at escalating prices. It started at $5000 with the HP 9100 and that served as a base price until after the HP 9825. Then the HP 9845 hit the scene and we found you could get machines costing $25,000 to $35,000 onto engineers desks. The $5000 machines didn't look so attractive any more. And that's exactly where the PC struck. For about 15 years after its introduction, several editors and analysts (including me) said that any PC you might want on your desk was about $5000, at any given time. By the mid 1990s, that number started to fall. HP had to start a new division that just built PCs to get back into the fray.
It was actually not the Desktop Computer Division, but the Data Terminal Division that came closest for HP to the $5000 computer. Beehive Medical Electronics, Alan Kay's first employer, was struggling to meet burgeoning volume and reliability requirements for Data Terminals that HP imposed, so Ed McCracken hired Jim Doub from Al Bagley’s F&T division where he had been building nuclear magnetic analyzers. Doub developed a superb Intelligent Terminals program, described in the HP Journal in June 1975. For HP, reliability was as important as screen editing, and their new line of display terminals exhibited high quality performance in customer environments.
HP Data Terminals, known for reliability and unusually nice interaction modes, quickly became big business. Orders for the first year were $17 million, doubling every year under General Manager Jim Arthur and R&D manager Jim Doub. Within five years, they exceeded HP 3000 CPU revenues. Peripherals were becoming central.
The HP 2640A was the flagship product at $2,640; in November 1975the HP 2644A was announced at $5,000, adding important storage capability to this interactive terminal. These operated very differently from a classic dumb terminal; because they had an internal microcomputer, they could function in their own right as a local computer, while also tethered to the CPU over a network connection. Terminals were gaining smarts, but also CPU independence.
HP was not alone in missing big signals, and DTD certainly was not alone among HP divisions in missing the personal computer revolution, but on reflection, DTD designers came closer to understanding this than anywhere else in the company – much closer than either the Corvallis team or the Ft. Collins team because they, as did IBM, were primarily dealing with business people, not engineers. And they had the price point right.
Wow, I was a summer intern 1978-1980 and walked around to see the HP 300, HP 250, and programmed the BASIC demo tape for the HP 2647 programmable graphics terminal, and witnessed the crazy color terminal they must have sold 10 of. Cut my teeth on the HP 9830 at my high school in 1975 which got me interested in HP. Seems HP 300 was a classic case of too many new ideas at one time, though many of the ideas are now widespread. I saw the soft keys evolve from paper labels to screen-locks on 2640 terminals to the HP 300 screen labels next to dedicated buttons that are on every gas pump and ATM today. Bit ironic that HP is now the biggest vendor of computers after all those blunders.
I remember Jim Doub. Learned how to drink beer at HP beer busts. Coffee and donuts were still the best I've had at any employer.
Ah, the beer busts! And the coffee and donuts were superb; not so good for a waistline as youthful designers began to age...
And HP was full of folk like Jim Arthur, who set one kind of tone for the place -- aggressive!
Post a Comment