Oral-History:Susan Bond

From ETHW

About Susan Bond

Susan Bond was born in 1942 and grew up in Essex, United Kingdom. After graduating with first-class honors from Bristol University, Bond joined the Civil Service in 1965. She was hired as a Scientific Officer for the Mathematics Division of the Royal Radar Establishment (RRE). Despite knowing nothing about computing at the time, Bond was placed on a nonnumerical computing research team. In this capacity, she worked on the RRE Automatic Computer (RREAC), Britain’s first solid-state computer. Just a few years later, Bond distinguished herself as one of the developers of the world’s first ALGOL 68 compiler. In recognition of her many accomplishments, she moved up the ranks at RRE over the next two decades. Among her various management roles, Bond acted as Computing Service Manager and Superintendent of Computing and Software Research. For most of her career, Bond was the most senior woman at RRE. She retired in 1993.

In this interview, Bond discusses her 28 year-long career at RRE. She outlines her transition from Scientific Officer to Superintendent of Computing and Software Research, describing each of the positions she held at the organization. Bond speaks at length about the principles and practices of compiler design, using examples from her own experiences, especially her work on the ALGOL 68 compiler. Moreover, Bond offers unique insight into the history of computing in the United Kingdom, particularly in the Civil Service.

About the Interview

SUSAN BOND: An Interview Conducted by Janet Abbate for the IEEE History Center, 26 September 2001.

Interview #579 for the IEEE History Center, The Institute of Electrical and Electronic Engineers, Inc.

Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

Susan Bond, an oral history conducted in 2001 by Janet Abbate, IEEE History Center, Piscataway, NJ, USA.

At the request of Susan Bond, there are additional restrictions on the use of this manuscript. Please contact the IEEE History Center for more information.

Interview

INTERVIEW: Susan Bond
INTERVIEWER: Janet Abbate
DATE: 26 September 2001
PLACE: Bond's home in Colwall, United Kingdom

[Notes courtesy of interviewer Janet Abbate]

Background and Education

Abbate:

For the record, this is an interview with Susan Bond on September 26th, 2001. We’re speaking in your living room in Colwall.

Bond:

Yes. And I’m trying to recall all the things that happened a long time ago!

Abbate:

I always start at the beginning, so: Can you tell me where you were born and where you grew up?

Bond:

I was born in 1942, and I grew up in Dagenham in Essex, and then I moved out later to Brentwood in Essex.

Abbate:

What did your parents do for a living?

Bond:

They were both teachers. My mother was an infant teacher and became an infant head teacher, and my father was a secondary school teacher, and he became an Audio-Visual Aids Advisor.

Abbate:

Did you have brothers and sisters?

Bond:

No. I’m an only child.

Abbate:

I guess if [your parents] were both teachers they probably valued education. Did they encourage you to go on and have a career?

Bond:

Oh, yes. This was the days of the eleven plus. I don’t know what you know about the English education system.

Abbate:

I’m not sure what the eleven plus is.

Bond:

It was a selective exam which you took at the age of eleven, which determined whether you went to grammar school or secondary modern school. It was optional whether one sat the eleven plus, and being in the teaching profession, they wouldn’t consider that I wouldn’t.

Abbate:

And the object was to go to the grammar school.

Bond:

I went to grammar school; that’s right. That was Brentwood County High School for Girls.

Abbate:

So you went to girls’ schools all the way through secondary school, probably?

Bond:

That’s right. From eleven till eighteen I was at Brentwood County High for Girls.

Abbate:

And were you particularly drawn to mathematics or science in school?

Bond:

Well, I suppose I always found that easiest, and got higher marks in that area. In England, in those days, you tended to do a broad span of studies up to O Level (which I took in 1959), and then you did two years’ study for A Level, for which you selected a smaller group of subjects. And so in the sixth form I did maths—maths, applied maths, and physics—for A Level, from which I then gained entry to university. I went to Bristol University from 1962 to 1965, and I read mathematics.

Abbate:

How did you choose that school?

Bond:

Bristol?

Abbate:

Yes.

Bond:

Well . . . I don’t know how much of this is relevant to your studies, but Bristol was said to be full of failed Oxbridge entrants: that is, it had a high proportion of people who’d failed to get into Oxford or Cambridge! [laughs]

Abbate:

People who had taken the exams for Oxford and Cambridge . . .

Bond:

. . . but hadn’t actually gained entry. Bristol was my first choice after Oxford or Cambridge. Very few people from our school actually went to either Oxford or Cambridge—perhaps one every three years, something like that.

Abbate:

But you had taken the exams for that?

Bond:

Yes, that was part of the third year in the sixth form, to take the Oxford and Cambridge entrance exams.

Abbate:

And was that common, in your school, for people to take them?

Bond:

No, very uncommon. I was born in the war, and so it was slightly before what was called the “postwar bulge” in 1945 when there was a tremendous surge in the birth rate, with high numbers in school classes. But still, our numbers were fairly high, and when I entered the sixth form it was the first year my school introduced a science sixth form.

Abbate:

Did you take the physics then?

Bond:

Yes. Up till then, the sixth form had been perhaps about twelve or fifteen girls; but in my year there were about forty, so they divided it into sixth form science and sixth form arts.

Abbate:

And before that it was just arts?

Bond:

Effectively, yes. There were perhaps one or two people who did science, but . . .

Abbate:

Did the corresponding boys’ school have science earlier?

Bond:

Well, there wasn’t a directly parallel boys’ school to ours. There were grammar schools in different towns, and some of them were single-sex and some—very few—were mixed.

Abbate:

So was it unusual to be doing the maths and physics exams?

Bond:

Yes: up till that time, yes.

Abbate:

It seems like it would have been remarkable to actually get into Oxford or Cambridge. You had to take a subject exam?

Bond:

That’s right, yes. There were seventeen students in upper-sixth science, and about three or four of us stayed on in the sixth form to take the Oxbridge entrance. But none of us went to Oxford or Cambridge.

Abbate:

I mean, considering it was the first year they even offered the sciences, it seems like it would have been remarkable if they’d got up to speed to actually getting people ready for sophisticated exams.

Bond:

Yes. That was the problem: they really didn’t have the staff to do it.

Abbate:

You went to Bristol, and . . .

Bond:

Yes, and I read mathematics. There were four women out of, I suppose, about twenty people reading mathematics; and at the end of the first year, all four women failed their exams!

Abbate:

How did the men do?

Bond:

Oh, well, they were all right. Perhaps one or two failed, but the majority passed. This was a Special Mathematics honors course.

And so we each had an interview with the professor, who put various options to us, and all four of us opted to change course slightly, and instead of reading special mathematics, to read for the general honors science degree, which was more varied. It took in some physics courses, and you could also do philosophy, and history of science. So it wasn’t quite such an abstract mathematical course.

Now, the reason that we failed, which we didn’t realize, was that all the men had actually covered most of the material for the first year when they’d been at school; so they arrived and the first year was absolutely plain sailing to them! But to us, the four women, it was all new.

So, we all transferred to the general science honors degree. And then, at the end of the three years, there were four first-class honors awarded—and the four of us got the four firsts! [laughs]

Abbate:

In the general science?

Bond:

In the general science, yes. After that first hitch, I thought it was extraordinary.

Abbate:

Very interesting.

Did you enjoy it there?

Bond:

Yes, I suppose I did; yes.

Getting a Job at RRE

Abbate:

Did you have some specific plans for what you were going to do with the degree? Or even general plans?

Bond:

Well, being an only child, I always thought, “Well I’ve got to be independent in my life. I’ve got to look after myself, basically.” And so I was interested in having a job where I could use mathematics; but I guess I wasn’t ever really interested in mathematics.

Abbate:

In being a professor and doing abstract research, you mean?

Bond:

That’s right. And because of not being in Special Mathematics but being in the General Honors, there wouldn’t be an opportunity to go into research anyway, even if I’d wanted to.

But I did want a secure job, thinking ahead to the days when I’d be on my own, so I applied to various places, including government establishments and also a couple of companies. This was in the very, very early days of computers.

Abbate:

Had you had any encounters with computers in college?

Bond:

No, none at all. I think, perhaps in my last year, they were just setting up the beginning of the Computing Center at Bristol. You know, it’s hard to remember how little there was. That was 1965 when I graduated.

Abbate:

Did you have any inkling that you might ever want to work with them?

Bond:

No; and in fact, when I look back at my job interviews, it amazes me what I said! [laughs] I would be happy to look at computers, but I wasn’t particularly interested in that; no. And actually, in my interview for RRE, where I finally ended up . . .

Abbate:

That’s the Royal Radar . . .

Bond:

The Royal Radar Establishment, that’s right. That was one of the places that I applied to; and because I’d got a first-class honors, I was eligible to apply as a Scientific Officer in research. In my recruitment interview, the chairman of the panel asked me if I was interested in doing research, and he tells me now I replied, “No!” [both laugh] He was the head of the Mathematics Division, and I subsequently got to know him very well, because he was my boss until he retired. I subsequently found that he was actively recruiting women, because he was very keen to build up the division in research, and it was difficult for the Civil Service to attract high-quality graduates at the time because of the salary, and he thought women might be a different pool—a different source for recruits.

Abbate:

Who was that?

Bond:

It was Philip Woodward, who had written a seminal book: Probability and Information Theory, with Applications to Radar[1]; and then became Superintendent of the Mathematics Division.

Abbate:

Well, did they want people who were going to do mathematical research? What were they looking for?

Bond:

The Mathematics Division was actually engaged in computing—and not numerical computing, which was the only thing I knew that computing could be, but operating systems and compiler-writing, because TRE had built its own computer.

Abbate:

And what does TRE mean?

Bond:

Telecommunications Research Establishment, which was the previous name of RRE—and they actually built a succession of computers. The first one, before I arrived, was TREAC; and then when I arrived it was RREAC: “RRE Automatic Computer.” That was a home-built machine, and had a 36-bit word and 24K words of core store.

Of course, when you build your own computer, you’ve got to program it right from the outset, so the research the Mathematics Division was doing was research into languages. They had written both the operating system for RREAC and an ALGOL 60 compiler, with compiling techniques. So the research team I was recruited for was, in fact, a non¬numerical computing research team.

Working on the RREAC

Abbate:

What was your first experience with—would this have been the TREAC? What computer did you use?

Bond:

The RREAC. RRE Automatic Computer.

Abbate:

What was that like to use?

Bond:

Well, it was what we’d call a mainframe now. It was in a room with an operator, and it was running a computing service for the Establishment. Users would come along, and they would do their own programming; but they could only use ALGOL 60, because that was the only compiler that was on the machine. They would type their programs up on five-hole paper tape. Then they would put them into a tray, and there was a chief operator who sat behind the console, which had rows of lights and switches—and she would solemnly take these boxes of users’ jobs, and she’d feed them into the five-hole tape reader, and then they would run, and when they were finished she’d take them out and go on to the next job. Since this was a computing service for the radar scientists—for the research scientists at the establishment—the research team who were trying to write compilers could not use it in the day; so if you wanted to do a debugging session which needed the console, you had to approach the operator on your bended knee early in the morning and say “Please, Ruth, can I use the machine for a quarter of an hour before you start the user jobs?” [laughs]

But I’ve skipped a lot of the process. You typed up your own program, and you rolled up your paper tape, and put it in the job box. There were banks of teleprinters in a specially sound-insulated room, and after the program had been run the operators fed the output tapes into the teleprinters and printed them off, then returned it to the users with their input tapes—their printout with their input tape.

Abbate:

How long was the turnaround?

Bond:

Well, it depended on how long the job was, and what program errors they’d made; but I suppose most of the jobs were five or ten minutes—except one or two much longer ones, and the aim was to complete the jobs submitted by the next day. One of the theoretical physicists, who became my husband towards the end of when I’d been there a couple of years, was the major user at the time; he had a program which was analyzing the dynamics of crystal lattice structure. That used to run all night, so the operators used to run it when they’d done everything else; they’d just leave it to use up the spare processing time.

So, it was a question of providing a service to all the scientists, who were doing radar problems and physics problems. The Mathematics Division used to help the users; we had to help them with programming, because they had to do all their own programming.

Learning and Using ALGOL 60

Abbate:

Was anyone using assembler at this point, or machine code? Or was it all in ALGOL 60?

Bond:

All the users were in ALGOL 60. We must have used assembler, but I don’t think I personally ever did very much. There was a feature in the ALGOL 60 compiler which said “code begin,” and then you were away—you could insert assembler code into the ALGOL 60 program, embedded in the framework of the ALGOL procedures. But the Mathematics Division had made a strategic decision that they were interested in easier ways of programming and more reliable ways of programming, so that’s why they went into high-level language.

Abbate:

What was ALGOL 60 like to use?

Bond:

Well, it was the first language I had used. I had to learn all this when I arrived; I knew nothing about computing whatsoever.

Abbate:

Was there a training course?

Bond:

I think we must have had them; yes. There was a book that Philip had written, called “Programming in ALGOL: Booklet” prepared by the Superintendent and Staff of the Mathematics Laboratory, Royal Radar Establishment published by the official publisher, Her Majesty’s Stationery Office (HMSO) in 1964. So I certainly read that. And the people in the Division did give the odd course; yes.

Abbate:

But it wasn’t hard to pick up for you, when you walked in without a background in computing?

Bond:

Oh, I was fascinated by this, because I had thought computing was numerical and number-crunching, and this work wasn’t. One of the early things that we had was a list processing package modeled on LISP. I never used LISP—it was all “CAR” and “CDR,” wasn’t it? — but RRE had written a package on top of the ALGOL 60 compiler which used the words “head” and “tail” instead of “CAR” and “CDR,” and so you’d write these programs, “hd(hd(hd(tl(thing)))) becomes such and such” [laughs]

Abbate:

But it came easily?

Bond:

Well, yes. It was basically even a level above using ALGOL 60 for programming. It was research that was trying to find out how to write compilers for high-level languages, and in the process one of the division, Michael Foster, designed and implemented a program called SID: Syntax Improving Device. That was one of my first projects, to work with Michael, because he was about to leave to take up a post at Aberdeen University—and obviously the Research Division didn’t want to lose his knowledge. So they said, “You work with Michael, and he will show you how SID works, and you can reimplement it, so that then you’ll know all about it!” So that was one of my first tasks, to reimplement SID in ALGOL 60 using the list processing package. [Much later I reimplemented it in ALGOL 68 using structures, which made the program much more transparent.] And what SID did was, it took in grammar rules for a language; it then transformed the rules of the language so as to generate an analyzer, which was effectively a reader [compiler compiler, see below] for the ALGOL or high-level language that you’d put the grammar rules in for.

Abbate:

Was that a lexical analyzer?

Bond:

No, it was at grammatical level, rather than lexical level. There is a paper on SID published in 1968, I think, in the British Computer Society Journal.[2]

Abbate:

Let me think about this: the ALGOL 60 specification was—I don’t know if its specification literally came out in 1960 or not . . .

Bond:

Well, it did more or less, yes.

On the Principles and Practice of Compiler Design

Abbate:

But it was only a few years old when you started using it in 1965. I wonder what kind of an art it was to think about compiler design: if there were well-established principles at that point, or was the point to figure out what is good practice in compiler design?

Bond:

Well, in a way, the language is designed with the compiler in mind. When designing a language, you’ve got to have some idea of how you’re going to compile it, and so therefore, the techniques you use . . .

[Susan Bond’s husband, Chris Sennett, enters the room]

Sennett:

Have you covered the topic of lunch yet? [laughter]

[RECORDING PAUSES FOR LUNCH BREAK]

Bond:

I guess I’ll collect my thoughts a bit. Yes, we were talking about compilers and languages, weren’t we?

Abbate:

Yes, I was interested in the practice of compiler design. So, I guess you’re saying in a way that when someone specifies a language, half of the work is maybe done—in the sense that if you specify it correctly, the design of the compiler follows logically. But how much leeway is there in designing the compiler, in terms of the kind of choices you need to make?

Bond:

Choices at what level?

Abbate:

Well, unfortunately I can’t speak from experience, so I’m not precisely sure; but presumably there must be decisions in the process of designing the compiler that would make it—I don’t know—more or less efficient, for example?

Bond:

Yes, that’s right. And not just the compiler, but there were always these debates about the object code being produced. There was a Ministry of Defense project in this country to design a high-level language for embedded computers; that was taken from some of the US Navy work on JOVIAL. It was called CORAL-64, and was a high-level language for an area which had always used assembly code—so the catch phrase was ‘the efficiency of the object code is of paramount importance.’ So the whole point of the compiler design was to make sure that a high-level language compiler generated efficient code rather than inefficient code. That was always a principle of our compiler-writing work. The object is to give the programmer high-level facilities for making more reliable programs that are easier to debug, but at the same time generating very tight object code: efficient object code.

I guess there’s the difference also in designing compilers for different sized computers. If you’ve only got small computers available to you, then you can’t spread yourself—this is before paging and virtual memory became common. It’s fixed memory allocation in lots of ways. And it also all ties in with the operating system and what features that’s got, and how it copes with multi-tasking, and the interrupt structure, and all that sort of thing. So it’s quite a large subject.

Abbate:

So you were actually writing pieces of an ALGOL 60 compiler?

Bond:

That’s right.

Abbate:

Was there a large team, each working a different piece of it? How did that work?

Bond:

There were about three of us, I think. The SID program was a . . . what’s the word? There was something called a “Compiler Compiler,” wasn’t there?

Abbate:

I used one called “yacc” on the Unix system . . .

Bond:

Oh, yes, yacc: “Yet Another Compiler Compiler.” That was the successor, wasn’t it?

Abbate:

So there was an earlier one?

Bond:

There was an earlier one called the Compiler Compiler [Brooker-Morris] at the time we’re talking about. So SID took in the grammar of the language and generated a program which would read in the code written in the high-level language and had function calls at the places where you would have to design how you were going to map from the language to the object code. It would analyze the input program: suppose it said “A becomes B.” SID would generate a program, which would be part of the compiler, which would read in “A becomes B,” and would do the lexical analysis, and then it would pass over the information that what you need to do at this point is to generate the code to do the assignment.

Abbate:

So it would specify all the requirements that would need to be—the things that actually would need to be done, and in what order?

Bond:

Yes, that’s it. That was syntactic, and the challenge then was to get as much semantics into the Compiler Compiler as you could—and that is pretty difficult. And it’s probably still being worked on to this day! [laughs]

Abbate:

I think it is!

But you’d obviously never studied compilers before you got there. I mean, there wasn’t . . .

Bond:

Well, it was all being done in research teams. It was only academic and US research teams, as I understood it. Well, I suppose the industry—who were commercially selling computers—obviously had to have their own compilers as well, didn’t they, and had to develop them?

Abbate:

Right. Well, I’m just thinking that there wasn’t a theoretical, pedagogical—it hadn’t really been developed to the point where you could go take a class in compiler design or something, like you would today.

Bond:

No, exactly.

Abbate:

So did you just sort of develop over time a sense of what would work? Or was there someone on the team who had already spent a lot of time thinking about principles of compiler design?

Bond:

Well, that was the team, really. After my first job reimplementing SID, I went on to work with Ian Currie, who was working on the CORAL project I mentioned before, to develop a compiler for embedded computers.

About that time, I think, it must have been that RREAC was showing its age, and RRE needed to replace it. There was no more building your own computers then, so they had to buy one. I’m going ahead a bit now, but I think about 1970, RRE bought an ICT 1907F for the computing service. We had the users using ALGOL 60 on our home-grown computer, RREAC, and that was all they could do; all it had, was ALGOL 60. The commercial computer, ICT, had FORTRAN. But the Mathematics Division, which was responsible for the computing service, felt that FORTRAN was not going to be adequate for the longer term for the users; it was difficult to debug, and all that sort of thing. So we started a project to develop an ALGOL compiler for the 1907F. By this time, the ALGOL 68 report was being written; so we actually decided that we would develop an ALGOL 68 compiler for the new computer.

To do that, we first bought a much smaller ICT computer called a 1902, which was the first one with an exchangeable disk store. We were years behind the States, I’m sure, in this, but that was what was commercially available. The exchangeable disk store was—how small was it? One megabyte, I think, perhaps? [laughs] So we were allowed to buy—and we had to make the case to the Treasury, because these things were all terribly expensive—to buy this ICT 1902 which had a 24-bit word, and 32K of main store. No floating point, of course.

Abbate:

Now, I’m not sure what you mean by “exchangeable.”

Bond:

Disk store. The backing store on RREAC was just tape: magnetic tape.

Abbate:

So this was the first magnetic disk . . .

Bond:

This was the first magnetic disk commercially available from a UK supplier. Because at the time the UK government policy was that any government establishment needing to buy a computer should go single-tender to a British company if it was more powerful than the Atlas computer at Manchester.

Abbate:

All right. Shall we take a break?

Bond:

Yes.

[RECORDING PAUSES]

On Women at RRE

Abbate:

Before we get to ALGOL 68, I meant to ask: Were there a lot of women in the computing group, besides you?

Bond:

No, I was the only one. I was the only graduate recruit. That was the thing that surprised me—I suppose because I hadn’t thought of it—when I arrived in Malvern.

Abbate:

The only woman graduate recruit, or at all?

Bond:

Woman graduate. At the time I arrived, there was myself and another recruit to the Physics Department, but she seemed to disappear within a couple of months. The Ministry of Aviation Supply was the ministry I joined, which was a military ministry, and there was a very distinct classification. There were Scientific Officers, who were honors graduates with an upper second or a first-class honors. And then there was the class called Experimental Officer, and that was for people with pass degrees, and Higher National Certificates, I think they were called at the time: people who had some qualification beyond secondary school but a notch down from the honors degree. I don’t recall any other women Scientific Officers; there certainly weren’t any working closely with me, until about three or four years later—and then, as I said, Philip Woodward was actively recruiting women, so our division was the one that tended to have women that he had recruited.

Abbate:

Now, there were only about ten people in your group?

Bond:

Yes. The whole division was twenty, when I joined, but that included the Computer Operators—who were all women, mainly from the local area. Ruth Hensman, who was the Chief Operator, was an Experimental Officer, who had joined the radar establishment in the war, and was called a “Computer” in those days. Did you know that people were called Computers? The Operators were Scientific Assistants, and they had to have Science O Levels, but not a very high qualification.

Abbate:

Was it possible to work your way up from one grade to another?

Bond:

Very difficult to move between the classes. You could go from Scientific Assistant but you had to get the right qualifications to carry on up as an Experimental Officer, and Senior Experimental Officer. And then there was the honors graduate class: Scientific Officer, Senior Scientific Officer, Principal Scientific Officer, Senior Principal Scientific Officer. That was the system.

Abbate:

I was wondering how far you had gotten in the . . .

Bond:

Yes, that’s what I was musing on. It was a very rigid promotion structure. It was partly to do with age and experience as well as ability, and it was an intricate system by which you were put up for promotion by your superiors to a certain grade. So, I joined as an SO. Now, a few years after, there would be an opportunity for promotion. There was an annual promotion review, and they would go through everybody who was a Scientific Officer and say, “Which of these should we put forward into the promotion review for Senior Scientific Officer?” If you were recommended to go into the review, then you had an interview—it was usually in London—from a panel of scientists, of which possibly none of them would actually be from your establishment, because this was a ministry-wide system. They would interview you to see if you were suitable to be a Senior Scientific Officer, and if they felt you were, then the panel would recommend that you were promoted. But because of the Civil Service structure and funding, there was a cutoff, in the sense that you couldn’t promote, perhaps, the number of people you might think were worthy of promotion every year. There was a limit, and this worked on age, combined with step potential. When you put a candidate up for promotion—this is somebody in your division—you had to give your judgment as to what grade that person was capable of reaching. What is their potential?

Abbate:

That’s step potential?

Bond:

That’s the step potential. So, when anybody’s being considered, say, for SSO promotion—you would say, “Right: This person is an SO now. What grade do they have the potential to reach? Would it be SSO, which is one step; PSO, which is two steps; SPSO, which is three steps; or higher?” You would come up to the panel with this recommendation—which you weren’t aware of yourself, as a candidate—and they would try and assess from the interview whether you indeed had the potential to reach the two steps, or the three steps, or whatever.

Abbate:

So you’d be more likely to be promoted one step if they thought you could eventually go the three steps?

Bond:

Yes, that’s right.

Abbate:

And did they have trouble imagining that the women were going to be very senior officers?

Bond:

Well—yes and no. In that scientific environment, most of them were open-minded. But not all; some were prejudiced. You could get discrimination by women not being put up for promotion in the first place. And of the candidates recommended for promotion—and I was on the panels in the end—you could get some of the panel members who would, I think, be prejudiced, but it never showed explicitly. Well, in my experience; I don’t know whether on all-male panels it would show more.

Abbate:

And were they just interested in your scientific ability, or were they thinking about leadership ability as well?

Bond:

Well, that depended. From Scientific Officer, three steps would take you to Senior Principal (SPSO). Now, what was called the “career grade” was two steps to PSO—Principal Scientific Officer. That meant that graduate recruits at SO would expect to reach the grade of PSO.

Above PSO, the posts were complemented: that is to say, there wasn’t a fluid number of SPSO’s. SPSO’s were attached to certain positions like Heads of Divisions, and Heads of research teams, and lots of posts in London on the admin side. So research scientists would expect, if they wanted to continue in research and not into management, to reach PSO; and then, for the real high-flyer research people, there was a system of Individual Merit Appointments, which meant that people could be promoted to SPSO grade or even higher on the basis of their published research papers and peer review, essentially. This system tended to operate in the Physics Department, but in Computing it was very difficult, because publications in computing at that time were fairly nonexistent. So, to get to PSO it would be the interview panel method, looking for potential, and for SPSO, you would either have to go the Individual Merit route, or you would have to apply for a job at the higher grade.

Abbate:

So there would have to be an open position as the head of a group, or a need for someone to head up a project, in order to create that job opening?

Bond:

Yes, it would be a permanent position. It would be something that the Treasury and the Ministry had agreed that these posts or jobs existed—and would be paid for, of course. [laughs] Because promotion was tied to pay, so you couldn’t go on promoting everybody, because the salary bill would be more than the Treasury would pay.

Abbate:

Was there equal pay for men and women in the Civil Service?

Bond:

Yes. In my career, yes. But the “potential” system was discriminatory; when I was put up for promotion to PSO, and wasn’t promoted, it was all to do with the assessment of my potential; I was told that somebody had said, in my line of management, “No, ridiculous; a woman could never reach that grade!”

Abbate:

So were there no female PSO’s?

Bond:

In RRE, I think while I was there I was the only one. And when I applied for and got the job of Superintendent of the Computing Division I was the first woman SPSO they’d ever had at RRE. I think there was somebody at Farnborough and there could have been some in classified projects, which you wouldn’t hear about.

Abbate:

But certainly, at the time they were thinking of promoting you, there weren’t a lot of other women already at that rank for them to say, “Well, of course a woman can do it.”

Bond:

No. That’s right.

Abbate:

Did you have any mentors or role models at work to advise you, encourage you, show you the ropes, that kind of thing?

Bond:

Philip Woodward: yes, Philip was my mentor, I guess. He who recruited me against all the odds [laughs]—against my disastrous comments at the job interview. He was Superintendent for about thirteen years, and then became an Individual Merit Deputy Chief Scientific Officer.

Developing the World’s First ALGOL 68 Compiler

Abbate:

Let’s go back to ALGOL 68, which we’ve abandoned.

I meant to ask: You bought a commercial machine for your group, and it came with FORTRAN. Was ALGOL not used? Manufacturers just didn’t provide it?

Bond:

That’s right.

Abbate:

Was it not widely used outside of a research context? Was the language not widely used outside of research groups?

Bond:

I don’t think it was, really. I think there weren’t all that many scientific programmers outside research groups at the time.

Abbate:

I mean, FORTRAN is a scientific language . . .

Bond:

Well, yes.

Abbate:

So if they were using a scientific language, it was FORTRAN?

Bond:

It was FORTRAN.

Abbate:

So that they wouldn’t feel the need for ALGOL?

Bond:

No. In fact, a lot of our users’ programs—which looked, as written for RREAC, as though they were ALGOL —I came to learn afterwards, when I found out more about FORTRAN programs and numerical libraries, that they were actually straight translations from FORTRAN to ALGOL! [laughs]

So in a way, we in the Research Group and the users were on different wavelengths, I suppose. But the Research Division felt that high-level languages were quicker to use and easier to debug, so that’s why we decided to provide an ALGOL compiler for the 1907F, and we decided that it wasn’t worth doing ALGOL 60 again, and this was when we got to ALGOL 68.

So we developed an ALGOL 68 compiler . . . Well, the first thing we did was to write what we called an Integer ALGOL Compiler, which was an ALGOL 60 compiler without floating point, without reals, because you don’t need them for compilers; and we called that IntAL. And then we used that language, IntAL, to write the compiler for ALGOL 68, to produce what was called the ALGOL 68-R Compiler for the 1900 Series.

The FORTRAN compiler for the 1900 series was very poor; I think it generated inefficient code, and had lots of drawbacks; and so when we produced the ALGOL 68-R compiler, it was actually a fast, efficient compiler for the 1900. And at the time, the government’s single-tender policy applied to universities as well, that if they wanted something over one Atlas power they had to buy British; and so a large number of the universities had 1900-series computers; and a lot of the departments fell on the ALGOL 68 compiler with cries of joy, because the only alternative was ICT 1900 FORTRAN, which wasn’t very good.

Abbate:

So they didn’t have much choice of the computer to buy, and once they’d bought it, there wasn’t a lot of choices for compilers.

Bond:

That’s right. There were two families of UK computers at the time: the ICT 1900 series and the English Electric System IV. We, of course, were in the 1900 community, and there were universities that we didn’t really have any dealings with, because they had System IV machines.

[CHANGE DISC]

Abbate:

Was anyone else writing ALGOL compilers?

Bond:

Ah, well, this is where we come to the famous book! [laughs]

Abbate:

[reading book cover] ALGOL 68 Implementation, by J. E. L. Peck.[3]

Bond:

ALGOL 68 was a very new language at the time, and this book is the proceedings of an IFIP Working Conference on ALGOL 68 Implementation, held in Munich, July 1970. I was working with two people, Ian Currie and John Morison, and I will reveal to you this photo! Have you seen this book, by the way?

Abbate:

No, I haven’t. [looking at photo] What a great picture!

Bond:

It’s been described as “a pop group of the early ‘50’s!” [laughs]

Abbate:

It does look like a singing group. [laughs]

So: there was a whole conference on . . .

Bond:

There was a conference on ALGOL 68 Implementation, and Ian wrote a paper; they asked for a photograph when they were publishing the proceedings. We weren’t really publishers in our research team. We spent all our time writing compilers, not writing about them. However, Ian produced this paper, and we all went to the conference; the three of us, plus Philip, as the Head of Division. In those days, you had to have special permission to go abroad; we had to go up several levels in the hierarchy for this visit to be approved.

Abbate:

Was that for financial reasons?

Bond:

Financial reasons. Yes, exactly.

Abbate:

They weren’t afraid you were going to leak state secrets.

Bond:

Well, no. This was always entirely unclassified.

The paper was accepted, and we all went to the conference. To say we made an impact wouldn’t be overstating it—because all these people were academics and in universities, and they had been defining this language, but nobody had been actually writing a compiler for it; so we found that when we turned up at this conference, we had the world’s first ALGOL 68 compiler—which absolutely thrilled the people who had written the language, because they hadn’t quite got that far. A different world from the commercial marketplace, of course, but it made quite a stir in those circles.

So we had the world’s first ALGOL 68 compiler, and because it was the best compiler on the 1900, and because there were large numbers of British universities with 1900s, we made it available, and it was distributed by ICL as free software.

Abbate:

Was that unusual?

Bond:

Yes, fairly. At that time, software was all the manufacturer’s software; everything was sold with the machine. But there was the ALGOL 68 compiler, and then there was the Numerical Algorithms Group at Oxford, who distributed a numerical library to all the universities. They weren’t confined to 1900s, though; they distributed versions for other machines as well.

Abbate:

Did this suddenly make ALGOL much more popular or widely used?

Bond:

Well, in the British universities with 1900 machines; and some of them actually took it up as a teaching compiler. What happened was that there weren’t any users’ manuals for ALGOL 68, so Philip wrote one and I was his co-author. That was the "ALGOL 68-R User’s Guide", describing how to program in ALGOL 68, and published by HMSO.[4] And it only cost 75 pence, so it could be a recommended text—it could be used with the compiler on the 1900s in the university departments.

Abbate:

So you could have students buy this.

Bond:

Yes, that’s right. So it sold 17,000 copies in the end.

Abbate:

17,000!

Bond:

Yes. Because it was a student book, and it went to two editions.

Abbate:

[Looking at book] It seems very clear and systematic.

Bond:

That one’s one pound twenty-five, but it’s still quite cheap for the time, and the students could afford to buy it.

Abbate:

So how did that change your life?

Bond:

Well, I ended up in computing support—supporting the ALGOL 68 compiler—because when people found bugs in the compiler, they reported them to me! This was all done within my job, and I did that for some years. We were still developing the compiler, and I won’t relate to you the horrors of it! [laughs]

Abbate:

Oh, do tell me a few!

Bond:

Well, it was all to do with maintainability, really; because the compiler, I’d told you, was written in integer ALGOL, and after a few years had gone on, we’d lost the ability to recompile the compiler. So when we had to correct a bug, we had to do binary patches, all through the store! [laughs] My part in the compiler-writing was to do the storage allocation, laying it out, and interfacing to the garbage collector. ALGOL 68-R had an in-built garbage collector; it didn’t just run out of store. It was all fully operational—because you need a garbage collector for ALGOL 68.

Abbate:

Isn’t that quite tricky, to make garbage collection work?

Bond:

Yes. But that was one of the strengths of the division, that they could do garbage collectors. Other implementations were not too bothered about running out of store. But our compilers always did garbage collection. And that’s why they were usable, because coming from the Establishment, the purpose in writing the compiler was not academic research into compiler-writing; it was to provide a language for the users to use. So it had to be usable, and it had to be efficient.

Abbate:

You could run a big program without running out of memory.

Bond:

Yes, that’s right.

Abbate:

In case you actually had some application that might need to use a lot of memory.

Bond:

Yes; —fascinating subject.

And then, on the strength of the "ALGOL 68-R User’s Guide" we were invited to give seminars at the universities which had the compiler, and I did quite a few of those. That must have been when I was SSO, and then, after some vicissitudes, I got promoted to PSO.

Becoming Superintendent of Computing and Software Research

Abbate:

When was that? Sometime in the ’70’s?

Bond:

It must have been in the ’70’s, yes. The first time I went for PSO promotion, the interview was an absolute disaster; and it was my fault—I wouldn’t have promoted myself if I’d been on the panel. So two years later, I was put up again, and this time I was more sensible and thought about the interview. It seemed that I’d done all right, but in the end, I wasn’t promoted; my step potential for my age wasn’t sufficient to get me above the line of the numbers who could be promoted that year. I was very disenchanted by this and made noises about leaving.

So I suppose the negotiating position was that I would go up for promotion again the following year. I succeeded in being put forward, and then I did get promoted to PSO. So, third time; but the first time was my fault.

Abbate:

I can’t remember how old you would have been at that point.

Bond:

I was 38.

Abbate:

Did you feel that you had to try harder to be taken seriously?

Bond:

Well, it’s better in the military services, because you’ve got a sort of badge. If you go to a meeting—and I found this particularly when I was Head of Division—you’re there as a certain position. I was involved in the NATO Defense Research Group Computing Panel, and was there as UK Delegate. So you are accepted.

At that time—and I’m just talking about the Defense Ministry now—there were far more women in the States than here. I was still the only one. I was the senior woman at RRE from probably when I got PSO until I retired. And I was slightly depressed, last year, when they put up a new building at the Establishment, and decided to call it the Woodward Building, after Philip Woodward. We were invited to the opening ceremony; and of the VIP guests, I think I was the only woman.

So I get a bit sorry about that—but then, your biology is against you, isn’t it? I had an incident with my boss, when I was asking why I wasn’t going to get promoted. “Well,” he said, “you won’t be here in two years’ time. You’ll leave to start a family.”

Abbate:

Was that a common assumption?

Bond:

Yes.

Abbate:

Was that an issue on your mind? How to balance work and family obligations?

Bond:

No, because I have never had any maternal urges! [laughs]

Abbate:

So that had nothing to do with needing to work full-time, or thinking it was incompatible with raising a family?

Bond:

No. I don’t want to get too personal with this, but I hadn’t honestly thought of even getting married, particularly. I was brought up to be very independent—because my mother was working, and so I was used to working women, and it never occurred to me that my life would be any different from that.

Abbate:

Did you ever have to take care of your parents, since you’re an only child?

Bond:

Ah, well, they’re 91 and 87 now, and one of the reasons I was quite happy to take early retirement was that I could see that I was going to have to look after them.

But the main reason I applied for early retirement was because I didn’t want the job that they were offering me. I was very happy as head of computing software research—the title changed through Software Research Division, to Software Engineering Division, and then to Systems Engineering Division. And that was fine, and I had thirty-five people in the division. And then we had a reorganization, and I had seventy people in the division.

Abbate:

Wow.

Bond:

And in the next reorganization, two years later, I was going to have a hundred and twenty in the division. I just didn’t want that job.

Abbate:

When you became Superintendent of Computing and Software Research—when you got that promotion—what did that put you in charge of?

Bond:

Well, two divisions were amalgamated. I inherited something that had been called the Computer Applications Division, which was working in embedded systems and also safety-critical systems; and then I had the Computer Software Research Division, which was the one I’d worked in, and that was software research, mainly compiler research. It had moved on from ALGOL 68-R to ALGOL 68-RS, which was a portable ALGOL 68 compiler; and then it moved on to the design of a generalized intermediate code after I became Superintendent. In the late ’80’s we got involved with the Open Software Foundation.

Abbate:

Ah.

Bond:

They did a Request for Technology for the Architecture Neutral Distribution Format—ANDF. The idea was that manufacturers could put software applications onto disk in this format, and then you could take it along to any machine—whatever the processor—and install it, because it was an intermediate code and would install on different architectures. The basis of it, of course, was compiler technology; and I had a team in my division that responded to the Request for Technology to the OSF. They were successful in being selected, and they ended up with a very fast C compiler technology, which would produce the ANDF. You would install this with one of a back-end set of installers which would go onto machines like 68K and the 86 Intel series. I think since I retired that’s been put into the public domain.

On the Relationship between RRE and the Ministry of Defense

Bond:

So I had the embedded systems people, the core of the old compiler team, and then a Computing Policy Section, responsible for input into the UK M.O.D. computing policy.

Abbate:

Ministry of Defense?

Bond:

Ministry of Defense. Of course, computing policy is a very difficult thing to set, and gets bound up with procurement policy; when the division expanded we had another section on requirements specification; all in the field of computing procurement.

Abbate:

That’s an interesting mix, then.

Bond:

It was, yes. That was when I got involved with the NATO Defense Research Group and also with a thirteen-nation IEPG program to develop a programming support environment. The US had the Ada APSE Initiative, and the Europeans had the PCTE program. It’s all initials, isn’t it? [laughs]

Abbate:

“Ada APSE” being Ada applications?

Bond:

APSE: Ada Programming Support Environment. The thought was that a programming system is not just a compiler, it’s all the tools as well, the maintenance tools. You need to procure these systems and keep them running reliably. But often the procurement process for a really big system goes on for twenty years, and when you look round you either haven’t got a system at all or it never works when it’s installed. All the failures of the 1980’s! Now it’s all changed, isn’t it?

Abbate:

I don’t know what’s going on now.

Bond:

Well, the military is becoming more orientated to the market. When I was about to leave, the US was working around to COTS—“Commercial Off-The-Shelf Software”—which is an attempt by the military to try and buy their systems from the commercial marketplace. I don’t think it’s a secret: in the [first] Gulf War, for example, none of the one-off systems really worked. Everybody would take their own laptops in!

Abbate:

[laughs] I never heard that.

Bond:

Oh, yes.

Abbate:

Right now the argument is, they keep trying to revive “Star Wars,” and all the computer people say, “Well, there’s no way this is going to work.”

Bond:

Yes. That’s right. David Parnas, when the “Star Wars” thing was first proposed, wrote an article about why computing systems should never be trusted—could never be trusted. But the other problems were to do with cost overruns and time overruns; and commissioning and building secure systems, which when they were put in were totally unusable, because all the security checks made them unusable.

Abbate:

So you had to be thinking about all of these procurement issues . . .

Bond:

Yes: my group. I was the manager of the research team, so selling the programs, getting the funding for the programs—because every year, the military would come up with savings measures, and you’d have to justify why you’re doing your research, and why they shouldn’t stop it tomorrow.

Abbate:

Now, who came up with the programs? Was it sort of bottom-up, that a research team would have a good idea? Or was it top-down, that there was a need expressed for this? Did you have input into what would be a good project?

Bond:

Well, that changed during my career. When I joined, I think it was pretty much bottom-up, but as the world developed, and we moved through the Ministry of Technology to the Procurement Executive, it became top-down. Our function as M.O.D. Procurement Executive was to support the procurement of M.O.D. computing systems, and at minimum, the research effort is there to train people, so that they can begin to grasp the problem at a higher level. So the research is not just a means to its own end.

Abbate:

Projects that you had, say, in embedded systems or something: Where would the idea for that come from?

Bond:

Well, people would be on call to advise procurement projects on embedded systems. So a big Army communication system, for example, or some of the really big M.O.D. projects, would hire research scientists to give them advice on the technical aspects and how they could get the requirement specification more honed.

Abbate:

So you were sort of consulting to ongoing outside things, where someone else in the military has decided they want to procure a system, and they just call you in?

Bond:

That’s right: they need an expert on computer systems, and my people could only keep their expertise up by having their own research interests, back at the ranch.

Abbate:

So it would be mix of doing original research, which presumably would someday be incorporated into Defense computing, and also consulting on here-and-now computer projects?

Bond:

Yes, that’s right. In fact, we had categories–and I think the US has a similar system: at the top level you’ve got basic research, which is your blue-sky stuff, and then you come back towards applied research, and then you come to direct project support. But in computing, you wouldn’t expect to apply your research software directly–when I started, you might, just might have expected the contractors to incorporate some of your work, but it was always a bit iffy; because the contractors have the responsibility, and they don’t want some technical lab’s product in the system that they’re tendering. So it became more and more that you had to get your research somehow out into the marketplace. You’ve got to somehow sell it out to the market, which is why we got involved with the OSF, because we could go back to our masters and say, “Look, we’re not just sitting here in Malvern amusing ourselves; we’re actually making an impact on the wider world, which will feed back into Defense policy.”

On Working with Universities

Abbate:

Did you also work closely with the universities? I know you said you had gone around giving seminars.

Bond:

No, we didn’t have all that close links to the universities—except during that particular period when we were supporting the ALGOL 68-R compiler, but that was very special.

Abbate:

So other than recruiting graduates to work for you, there wasn’t a lot of exchange of ideas?

Bond:

No, [although we funded some university research programmes].

On Commercial Applications of Research

Abbate:

Did you ever spin things off to the private sector?

Bond:

Well, we tried. There were lots of government initiatives to encourage the Establishment to do that. One of the big successes of RRE was liquid crystals; the RRE Physics Department holds several worldwide patents. And now they’ve changed the Establishment into QinetiQ, which is privatized, the royalties from these patents are very important. But in computing, tell me: how do you make money from software? [laughs]

Abbate:

Ask Bill Gates.

Bond:

Yes, well exactly! But ask all the others who have disappeared. There’s that book called Accidental Empires that’s written about Bill Gates and Steve Jobs and all those people.[5] I am fascinated by that.

On Being a Woman Manager

Abbate:

What accomplishments are you most proud from your time as Superintendent?

Bond:

It was the OSF thing, I guess. I was very proud of my Division. That’s probably a female attribute; I don’t suppose men feel the same. But they were really good—at the end I had five or six Individual Merit Scientists in the Software Division.

And what I learned from Philip was that the art was to protect your people and let them get on with producing something in their research, while staving off the attentions of the bean counters! [laughs]

Abbate:

Did you find that the management aspect came easily?

Bond:

Well, I think it did when I started; but the world changed so much. And the overall management of the establishment was always, as one of my colleagues said to me, ten years behind current management theory, so you were against the prevailing culture. But I was very focused on keeping in touch with the work, and I thought part of the function of the Superintendent was to act as a sounding board. Not in detail, but to be able to at least hold a conversation with your people–the research leaders–and perhaps divine where they were going; perhaps influence it a bit. Then the job became more and more staring at spreadsheets, and explaining why your division had spent thirty hours last week doing X, Y, and Zed, and why you were over budget because X, Y, and Zed had happened—and I just wasn’t interested in that. For example, in the old days, you published technical memos–RRE Technical Memos–and the Superintendent had to sign them off. And I always liked to read them all; they’d put them to me, and I would tell them what I thought was wrong with it! [laughs] Well, you know what I mean: just try and help it to come over more clearly. But towards the end, there wasn’t time for that. As the division got bigger and bigger, you just couldn’t do it.

Abbate:

What have you found to be the most satisfying aspects of working with computers? It sounds like you’ve had a lot of fun with it.

Bond:

Yes, that’s right. I don’t know if this is the sort of answer you’re looking for, but what I enjoyed most was debugging. I like debugging, because it’s interesting to get the diagnostic process going, and to try and work out why these interactions are happening. I certainly spent many happy hours on the floor poring over compiler and system dumps. [A system dump was printed on continuous line printer paper about 16 inches wide, and from half an inch to 2 inches thick; so it was easiest to spread it out on the floor to read. Everything was printed at that time; there were no monitors. Each time the system crashed we would decide whether the problem was hardware and the on-site engineers should be called in, or a software problem to be reported to ICL.]

Oh, I’ve missed out completely one job that I did: at the time I got PSO, Chris was Head of the Computing Service; so he was in charge of the 1907F, which at that time came up for replacement, and at the same time he was heavily involved in the George III operating system (the 1900 operating system, based on Multics) and extending it for a major Defense Scientific Computing Network project, and it was all too much; so I was asked by my Superintendent at the time if I would take over from him as head of the Computing Service. So that was my stepping stone, really, to being the Superintendent—because I ran the mainframe service for three or four years.

Abbate:

So between being PSO and Superintendent . . .

Bond:

That’s right: just shortly after I became PSO, I moved on to be Computing Service Manager, and we upgraded the 1907F to a 1906S, which was the very last 1900—and boy! We got 512K of plated-wire store! [laughs] I stayed with the Computing Service until it was coming up to that machine needing replacing. But I didn’t want to stay there very long, because I didn’t want to get tarred with the sort of female job of running the Computing Service—people can accept women managers of facilities perhaps more easily than they can accept women managers of research teams.

Abbate:

Have you found that women tend to end up in certain areas of computing? Either content areas or types of jobs? I know you mentioned running computer centers . . .

Bond:

I couldn’t really say.

Abbate:

Maybe there wasn’t a big enough sample of women at the time . . .

Bond:

Well, that’s right. I was thinking that, yes.

Abbate:

Now, I haven’t found a huge number of women in operating systems . . .

Bond:

No. Have you talked to the George III people, because there were quite a few women on the George III Development Team? Many of them were working from home as well. They were pioneers in setting that up. That was run from the ICL headquarters at Bracknell.

Languages: well, language development—there were not all that many people doing language development, were there, really? Either in the States or here.

Abbate:

Well, I guess in the ‘50’s there were. . .

Bond:

Well, there’s Grace Hopper, of course! I was on the Ada Board, as part of my job, for a couple of years, and there was Jean Sammet. She’s written a history, hasn’t she, of computing languages? And then Virginia Castor was the Head of the Ada Joint Program Office; but I think she was possibly a Defense career engineer.

Abbate:

I’m going to ask some general questions, unless I’ve missed something. I don’t know if I skipped any pieces of your career.

Bond:

We just filled in that gap that I completely forgot about the Computing Service.

On Changes in Computing

Abbate:

What strikes you the most about the way the field of computing has changed since you started? I mean, obviously it’s changed a lot; I’m just curious about what interests you the most about the way it’s changed.

Bond:

Well, I retired in 1993, so that’s the end-point pinned.

My first computer was RREAC: you walk in to the room, and you’ve got a big cabinet here, and a big console there, and a minuscule amount of core store. And now I’ve got in there [points to PC], I’ve got a Celeron processor, a hundred and twenty-eight MB of RAM, and a six-gigabyte hard disk—and an operating system I don’t know anything about! [laughs]

Abbate:

Do you think the culture of computing has changed? I mean the kind of people who work in computing; the public image of computing . . .

Bond:

Well, I assume that most of the software writers these days don’t understand how the machine works. They don’t think in the same way as we did; we had to understand. When I started, I wanted to know how the disk worked and how it stored the information, and I realized that you’ve got to get the inputs to the processor, and all that sort of thing; but they don’t have to think about that anymore. They can just sort of spread themselves and develop Visual Basic and C++ programs. I don’t know anything about how they debug them, or document them, or anything like that.

I think that’s one difference. There’s also the gulf between mainframe development and what’s become PC development. Businesses have got to use reliable operating systems, and I think a lot of businesses are still tied into the old mainframe manufacturers, albeit under different names. I met somebody last week who still maintains a VAX system at the Establishment, which is all on Alpha processors, and he said he’s not quite sure who he works for: yesterday he was working for Compaq, but perhaps today he’s working for HP!

Abbate:

It was DEC originally, right?

Bond:

Yes, exactly. And he thinks that as long as he’s around, they will keep the system, because it’s running perfectly well and it doesn’t go down more than once a year. But when he’s retired, they will probably junk the machine as well! But there’s nothing wrong with it.

Abbate:

No one knows how to maintain it.

Bond:

You will understand, having written a book on the Internet: it’s to do with the difference between fluidity and things sort of happening by synergism, as against more rigid and controlled development.

On the Status of Women in Computing

Abbate:

Do you think the computing field has become more open to women over time, or less open?

Bond:

I can’t say. I read that in the UK the number of women applying for undergraduate computing courses is falling.

Abbate:

Do you have any idea why?

Bond:

Well, there could be a perception that it is a sort of soft subject. But perhaps that’s my prejudice coming out: ‘Is computing science?’ Because at the undergraduate level you’re not doing any abstract, high-level thought, are you? You’re doing programming. How you teach computing at graduate level I don’t know. Because none of us was taught—none of my contemporaries.

Abbate:

Well, there weren’t—until the mid-sixties, I don’t think there were any computing programs.

Bond:

That’s right.

However, I think I’ve rambled on enough.

Abbate:

One more question?

Bond:

Yes.

Abbate:

Do you have any advice for young women who might be contemplating a career in computing?

Bond:

No! [laughs] I cannot give them advice from my advanced age. The world is so different.

Abbate:

Do you think there are any constants in computing? In terms of practice in the field, or what might be good aptitudes, or what’s enjoyable about it? I don’t know: anything?

Bond:

No. Sorry, I really cannot give any sensible answer whatsoever.

Reflecting on Her Career

Abbate:

If you had it to do over again, would you do it again?

Bond:

Well, if I were back in 1965 and had my career over again, I’d be quite happy, yes. I wouldn’t have changed anything. I was very lucky. It was amazing.

Abbate:

What do you mean?

Bond:

Well, I had a wonderful career. I enjoyed it.

Abbate:

Was the technology amazing? Did it feel like you were on the cutting edge all the time?

Bond:

In software, oh yes! The OSF-RFT was an indication. But it’s getting the cutting edge translated into use, isn’t it? It’s the gulf between the research and the market, and the market’s coming up from a different place. Accidental Empires—how the PC and current computing have developed—it’s not predictable, or even plannable, a lot of it’s just sort of accidental.

However, I’ve philosophized too much!

Abbate:

Well, thank you very much!

Bond:

Well, thank you.

Notes

1. P.M. Woodward, Probability and Information Theory, with Applications to Radar (New York: McGraw-Hill, 1953).

2. J.M. Foster, “Syntax Improving Device (SID).” Computer Journal 11, no. 31 (1968).

3. J.E.L. Peck, ed., ALGOL 68 Implementation, Proceedings of the IPIP Working Conference on ALGOL 68 Implementation, Munich, 20–24 July 1970.

4. P.M. Woodward and S.G. Bond, "ALGOL 68-R Users Guide" (Malvern, England: Royal Radar Establishment, 1975).

5. Robert X. Cringley, Accidental Empires: How the Boys of Silicon Valley Make Their Millions, Battle Foreign Competition, and Still Can’t Get a Date (Reading, MA: Addison-Wesley, 1992).