Oral History of Webb McKinney Interviewed by: Michael Mahoney Recorded: June 6, 2008 Mountain View, California CHM Reference number: X4701.2008 © 2008 Computer History Museum Table of Contents DIGITAL INSTRUMENTATION AT HEWLETT PACKARD ..........................................................4 R&D FUNCTIONS FOR DIGITAL INSTRUMENT COMPUTING..................................................6 OPERATING SYSTEMS.............................................................................................................10 LESSONS IN DEVELOPING SOFTWARE.................................................................................12 MULTIPLE MICROCOMPUTER PROJECTS AT HP .................................................................13 PERSONAL COMPUTER SOFTWARE......................................................................................15 COMPETING TECHNOLOGIES.................................................................................................18 HP IN THE PC MARKET ............................................................................................................20 R&D MANAGEMENT IN THE UK ...............................................................................................21 NEWWAVE .................................................................................................................................24 THE CHANGING PC MARKET...................................................................................................25 ANOTHER SHOT AT THE PC CORPORATE MARKET ............................................................26 CHM Ref: X4701.2008 © 2008 Computer History Museum Page 2 of 30 Webb McKinney Conducted by Software History Center—Oral History Project Abstract: Webb McKinney was an executive in Hewlett Packard over many years and involved in various efforts to build software to run on HP instruments, specialized equipment and eventually on minicomputers and personal computers. His recollections illustrate how the decentralized operation of many of the HP products led to multiple developments of both minicomputers and personal computers with conflicting architectures and business models. While these competing technologies would get resolved over time, HP was often late to market and never really focused its computer software on the broad personal usage market. HP wanted to differentiate its hardware platforms through its unique software, but the market migrated to a Microsoft Windows platform and neither the customers nor the software producers were prepared to go a separate route. Michael Mahoney: I'm Michael Mahoney and I am interviewing Harry Webb McKinney of Hewlett Packard. We are at the Computer History Museum in Mountain View, California and it is June 6, 2008. This interview is being done as part of the Software Industry Special Interest Group’s Oral History Project. Tell me how you got to Hewlett Packard. Webb McKinney: Well, actually, it's an interesting story. My mother had a very close friend who moved from southern California to Los Altos, California, which is very close to Palo Alto, which is where HP is located and we used to visit her in the summers for vacation and from the time I was eight, I wanted to live in the Bay area. And so, when it came time for me to graduate from the University of Southern California, although most of the people went into the aerospace business in southern California, I wanted the Bay area. I'd had a summer job in aerospace and I didn't like the environment, the sort of hire and fire environment there. And so I asked my uncle, who was a stockbroker, what were some good companies in the Bay area and he said well, I hear great things about Hewlett Packard. So I interviewed with them and ended up here. Mahoney: Were you a double EE major? McKinney: Correct, Yes. Mahoney: You got a Bachelor and a Masters at Southern Cal? CHM Ref: X4701.2008 © 2008 Computer History Museum Page 3 of 30 McKinney: Both, Yes. Digital Instrumentation at Hewlett Packard Mahoney: In what capacity did you come to Hewlett Packard? McKinney: Well in those days they asked you whether you were into analog or digital and I said digital, so I went to a division that was called, "Frequency and Time," which subsequently became the Santa Clara Division and I worked on probably the first major fully digital instrument. This was in the instrumentation part of the company, which is now Agilent Technologies. It's a separate company from Hewlett Packard, but it was in instruments and this was the very beginning of the idea that you could digitize the wave form. You do everything with digital computers and then if you need an analog output, you make it into analog signal again, using DA [Digital-to-Analog] at the end. It was all pretty sophisticated for those days: array arithmetic, Fourier transforms, modal analysis, a lot of math basically to do the measuring. So that was how I ended up in that organization. Mahoney: Were you working on the products themselves? McKinney: Yes, actually my first year was in marketing. The product line that was called Fourier Analyzers at the time had just come to market in I think this was 1969. The Fast Fourier Transform, the Cooley-Tukey algorithm, had just made it practical for computers of that era to actually compute Fourier transforms fast enough that they could be used for spectral analysis, and digitizing in those days was sort of a 25 kilohertz useful bandwidth, so the applications tended to be mechanical in nature, measuring structural resonances of cars, planes, trains, helicopters, buildings. They had just started coming out with this product, which was totally new for HP, both the technology and the application. I was on the marketing team and I trained the sales force representatives. I trained some of the first customers and then I moved into R&D where I spent the next 11 years or so in R&D on that product, developing the product, and subsequently being the R&D Manager for the Fourier Analyzers and a few other products. Mahoney: How were you identifying who your customers were when you were marketing it and what was your relationship to them? McKinney: Well it was really, “who does things below 25 kilohertz?” And of course HP, at the time, the main instrument sales force was primarily selling to electronic designers, for microwave design; this type of thing was really where most of it was. We actually had to create our own sales force within the overall instrument sales force which was focused on mechanical applications. And we were coupled together with another product line call "Laser Interferometry," which made very precise distance measurements that was, at the time, used in machine tools and now is, by the way, built into almost all the wafer stepper and IC CHM Ref: X4701.2008 © 2008 Computer History Museum Page 4 of 30 manufacturing equipment; but in those days, it was a mechanical application, so they were both from the same division. We said let's put lasers and Fourier analyzers together and create sales specialists out there. So in the beginning, it was a question of some prospecting and trying to find who really is interested in this. We pretty quickly found quite a few customers in transportation for people designing vehicles, space vehicles as well as automobiles. Mahoney: To what extent did this involve educating them in what they needed? McKinney: Well it was a different way for them to solve their problem. In fact, in those days, I have to say that the Fourier Analyzer computer that we used was a minicomputer, so it was a 2114, 15 or 16. You had a choice. I remember that the main product had 16K bytes of memory, so you put that in the context of probably what's in your wrist watch today, but it was a rack full of equipment at the time. I think that this was relatively new for the customers in terms of the way to make these measurements. They hadn't really been able to do this before. I used to say to get a Fourier Analyzer you needed $100,000 and a PhD, because the customers that actually used this equipment typically were PhDs. They were very sophisticated, highly educated people. This ultimately became a much broader market and we came out with much lower cost, simpler to use products. But in the very beginning you could almost think of this as a giant programmable array calculator and people had to know a lot about not only the problem they were trying to solve in terms of their structures, but quite a bit about math and measurement in order to actually get a result out of one of these. Mahoney: This actually gets us towards software. McKinney: Yes. Mahoney: Who provided the software for that? McKinney: Well, we did it all. Mahoney: Did you sell a package? McKinney: Over time, we found that there were general purpose sales, so we would sell one of these $100,000 systems and not be too sure what it was used for. We also found, over time, that there were some specific applications where we developed software beyond the norm. So the software that was in the Fourier Analyzer was really the software that allowed customers to write a program in a high level language to say, “Okay, take from the analog to digital converter, put it into block one, take a Fourier Transform of it, square it, add it to that.” So it was kind of like a big calculator program, except instead of operating on individual numbers, we were operating on arrays and so we did all the software that allowed people to write those programs. And then over time we developed some specific programs, one was called "Modal Analysis," which CHM Ref: X4701.2008 © 2008 Computer History Museum Page 5 of 30 actually showed you how the first mode of vibration would look and/or the second mode. In fact, one of the famous applications after the fact was for the Lockheed Electra, which, you probably remember, had a problem where the propellers excited a vibrational mode in the wings and the wings would fall off. It's like the famous Tacoma Narrows Bridge. Modal analysis is a sophisticated technology that allows you to actually measure those modes of vibration and then display them to show kind of what they look like. So we did a fairly sophisticated piece of software and sold it as a modal analyzer. There was the general purpose Fourier Analyzer.
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages30 Page
-
File Size-