IEEE

Oral-History:Lois Haibt

SHARE |

From GHN

(Difference between revisions)
Jump to: navigation, search
Line 1: Line 1:
 
==About Lois Haibt ==
 
==About Lois Haibt ==
  
Lois Haibt is a computer scientist who worked at IBM developing FORTRAN. She was the only woman member of the FORTRAN team at IBM. She joined IBM soon after graduating from Vassar, where her interests were in mathematics and science. She worked on several other computer programming projects while at IBM after FORTRAN.  
+
Lois Haibt is a computer scientist who worked at IBM developing [[FORTRAN]]. She was the only woman member of the FORTRAN team at IBM. She joined IBM soon after graduating from Vassar, where her interests were in mathematics and science. She worked on several other computer programming projects while at IBM after FORTRAN.  
  
 
In this interview, Haibt talks about working as a computer programmer for IBM out of college on the 704 computer. She also talks in depth about her work on the FORTRAN compiler. Haibt also describes her experience as one of the few woman computer programmers in the field during her start at IBM in the 1950s and onward.  
 
In this interview, Haibt talks about working as a computer programmer for IBM out of college on the 704 computer. She also talks in depth about her work on the FORTRAN compiler. Haibt also describes her experience as one of the few woman computer programmers in the field during her start at IBM in the 1950s and onward.  

Revision as of 17:42, 24 March 2014

Contents

About Lois Haibt

Lois Haibt is a computer scientist who worked at IBM developing FORTRAN. She was the only woman member of the FORTRAN team at IBM. She joined IBM soon after graduating from Vassar, where her interests were in mathematics and science. She worked on several other computer programming projects while at IBM after FORTRAN.

In this interview, Haibt talks about working as a computer programmer for IBM out of college on the 704 computer. She also talks in depth about her work on the FORTRAN compiler. Haibt also describes her experience as one of the few woman computer programmers in the field during her start at IBM in the 1950s and onward.

About the Interview

LOIS HAIBT: An Interview Conducted by Janet Abbate for the IEEE History Center, 2 August 2001

Interview # 598 for the IEEE History Center, The Institute of Electrical and Electronics 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, 39 Union Street, New Brunswick, NJ 08901-8538 USA. 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:

Lois Haibt, an oral history conducted in 2001 by Janet Abbate, IEEE History Center, New Brunswick, NJ, USA.

Interview

INTERVIEW: Lois Haibt
INTERVIEWER: Janet Abbate
DATE: 2 August 2001
PLACE: Caldwell, NJ

Background and Education

Abbate:

This is an interview with Lois Haibt, on August 2, 2001.

To start out, could you tell me when you were born and where you grew up?

Haibt:

I was born in 1934 and grew up in Lower Westchester. My family moved East when I was three, and I guess I feel like I’ve lived in Westchester all my life.

Abbate:

You moved East from . . .

Haibt:

From Chicago.

Abbate:

What did your parents do?

Haibt:

My father was an optical engineer. My mother was a housewife.

Abbate:

Did you have brothers and sisters?

Haibt:

I have one older brother, who is also technical. He went to Annapolis so he could play football, and then he became a pilot and stayed in the Navy so he could continue to fly. He was just a few years younger than all the pilots that were in the Korean War, so if he had tried to get a commercial job, he would have been a co-pilot for all his career, because of the seniority. There was this big bulge of pilots just a couple of years older than he was that came out of the Korean War. So he stayed in the Navy because he wanted to fly.

Abbate:

Did your parents encourage you to have a career, or to work?

Haibt:

Not explicitly, but they always encouraged me academically—and I was always a very one-sided math and science person.

Abbate:

So, from an early age . . . ?

Haibt:

From an early age—which made it sort of hard, in those days. Nobody was supposed to be good in math, and particularly not girls. Girls were good in languages; in writing and poetry! Even more than the “geeks” nowadays . . . So it was always rough, because I tended to be in classes with mostly boys.

Abbate:

You were going to regular public schools?

Haibt:

I went to public schools. They were very good ones, fortunately.

I went to college at Vassar because they had a very good science department and they had a very good scholarship program (which I needed badly at that time, because my father had been ill a while), and I just decided I’d had enough hassle. I thought at the time that I probably could get into MIT if I wanted to, but I just couldn’t imagine putting up with that.

Abbate:

Really? Because it would be a hostile environment?

Haibt:

Yes. I mean, it would be me and all of fifty-five zillion young men!

Abbate:

I’m sure it would be.

Haibt:

In those days, even more than now, they didn’t always take kindly to a girl showing them up—which I did regularly in high school. I even got a lecture from a math teacher one time, that I shouldn’t be so gleeful when I could do something that they couldn’t do.

Abbate:

Really!

Haibt:

It was a woman, too, which surprised me.

Abbate:

Would she have given a boy a lecture on that if he’d done the same thing?

Haibt:

I don’t think so.

Abbate:

So this was specifically: “It’s not ladylike.”

Haibt:

Right. Because she had put an extra-credit problem on the board, and a couple of other people had tried to solve it, and I’m saying, “I know how to do it!” [laughs.] That’s when she told me not to be so eager.

Abbate:

So you went to Vassar and majored in math?

Haibt:

Yes. I got through college not quite knowing what I was going to do, because I had known I never wanted to teach. I was extremely shy. (I’m still extremely shy, but not like I used to be.) I knew I didn’t want to teach. At that time many of my classmates ended up going to Katherine Gibbs for a year so they could get jobs as secretaries, or they got jobs as teachers—because that was a period in which teachers were in very short supply, so that even though they’d had no teaching courses they could get a provisional job, provided they went to summer school for the next couple of years to get the right number of education credits in addition to the bachelors’. So a lot of my friends in college, classmates, got jobs as teachers.

Abbate:

But you didn’t want to do that.

Haibt:

I didn’t want to teach, did not want to stand up in front of a group of people.

[recording pauses]

Working for IBM

Abbate:

So: You were finishing up and didn’t really know what you were going to do. Had you had any summer jobs or anything?

Haibt:

I’d worked at Bell Labs for a couple of summers, which I liked very much. Before that I’d gotten work locally in Lowell Manufacturing Company, in the office, in high school. There was engineering, but that was a really tough field for a girl in those days—probably still tough, but not like it was then. That was 99.999 percent male! So I wasn’t entirely sure, and I was concentrating on what I was doing. Then an interviewer from IBM came to visit what we called the “Vocational Bureau” at that time; placement. I had an interview at college, and then they had me come to New York, and when I got through they offered me a very nice job and a very nice salary—which was, I think, at that time the highest anybody at Vassar had ever been offered.

Abbate:

Do you remember what that was?

Haibt:

Yes. Fifty-one hundred dollars.

Abbate:

And this was what year?

Haibt:

‘55. More typical was $3000, at the time.

By then I had known a little bit—I didn’t know anything about computers, but I knew people who worked in computers. Because IBM is in Poughkeepsie [where Vassar is located], and they hired a lot of young [male] engineers straight out of college—and guess how long it took them to find the college? [laughs.] So I had met some people—in fact, eventually one of them I married; but this was before I started dating him. I had heard them talking about computers, and didn’t quite know what it was all about, but . . .

Abbate:

So you had never used one before you worked for IBM?

Haibt:

No. I mean, they were at the ten-in-the-world level in those days! [laughs.] You know, “What’s the forecast for this?” “Well, maybe the whole world could use ten of these.” Things have changed! [laughs.] You know, there was nothing like personal computers. I mean, the original Radio Shack TRS-80 was very comparable in amount of memory, speed, power—not physical size, but power—to the IBM 704 I started out with: almost exactly the same cycle time, memory size, and so forth. However, in those days, it was a multi-million dollar machine that took a large, heavily air-conditioned room that looked like a large appliance showroom—with all-gray appliances! [laughs.] It looked like refrigerators and stoves.

IBM was going to pay me a lot more than Bell Labs offered. Also, it was right in mid-town [Manhattan], and Bell Labs at that time was down at West Street, what’s now the Westbeth artists’ complex at 12th Street and the Hudson River. [To get to Bell Labs] you took the train in, and then the shuttle—I think you had to take two trains because of the express stops—and then you had a fifteen-minute walk; and I had had pneumonia my senior year in college, so that was bad enough for a summer job, but picturing going there in the winter!

Abbate:

So the commute really was the final factor?

Haibt:

Well, it was the commute; the fact that I was getting paid twice as much; and also, at Bell Labs I was offered a job doing the programming for the mathematicians—in other words, being an assistant—and at IBM, I was going to be doing things.

Abbate:

What were they offering you at IBM?

Haibt:

Well, I wasn’t completely sure, but they gave me a math test, in addition to the programmers’ aptitude test. They had a bunch of questions that that department had [put together], which were talking about differential equations and whatever—which they explained I didn’t have to worry about, because this was what they gave to the Ph.D.s. I knew the answers to all of them with no problems (except one, which I knew was partial differential equations, which I knew enough to recognize that that was a graduate course). So I figured, if they were going to ask me that kind of question and pay me that kind of money, they were not going to ask me to type somebody else’s result! They were going to expect a little bit more. And I knew the people that were doing computing things loved it. So I wasn’t completely blind, but I didn’t really know a lot about it.

Abbate:

What was the first thing you did there?

Haibt:

Well, the first thing was that I was put in with all the other new hires in a training group. We were sent to a programming class for a couple of weeks, and given some test problems.

Abbate:

And that would have been assembler at that point?

Haibt:

Oh, yes! Compilers didn’t exist. Assemblers were fairly new—just getting off the primitive stage. By today’s standards they were very primitive, but at least they had assembler.

Abbate:

And this was for the IBM 704?

Haibt:

704. I started in the first of July, and the first 704 was not delivered to IBM New York until December, just about Christmas time. It was running somewhat after Christmas.

Abbate:

This was ‘55?

Haibt:

Yes. We had the 701 before that, which I did not learn because they started us on the new machine.

Abbate:

So you showed up with no computer to work on?

Haibt:

That’s right. You know, they had me doing some training programs; everybody was getting ready for it.

Your program was a much slower thing in those days. You didn’t type in a few things on your terminal and press a button; you laboriously wrote out rather detailed instructions, and then you had to get them key-punched, and then checked, and then eventually you assembled them, and you checked that; and you ran the machine yourself. You went down and got a half hour of time or something.

Abbate:

You did your own operating?

Haibt:

We did everything; oh yes. The nearest they had to an operator was what they called “Mathematician of the Day.” That was the title. They didn’t even have a “Programmer” title; it was “Mathematician.”

Abbate:

That’s what your title was?

Haibt:

It was Assistant or Associate Mathematician. I don’t remember the qualifier anymore.

They would take turns spending the day, or some time, down on the machine, in case anybody needed assistance; but most of the time that meant sitting over on the side doing whatever. But programming was much more a pencil-and-paper operation, because machine time—even when we had the machine—was so scarce. You did a lot more by hand. Now you can just punch a button, see what happens! [laughs.] [But back] when you scheduled time days—if not a week—in advance, you didn’t go down and see what happens; you made pretty sure it was [correct].

Abbate:

So you’d try to check it in advance, to make sure it was right?

Haibt:

Yes, you tried to do the best you could. And [you’d] look at all the side-results and the intermediate things to check everything you could when you did get a run, before you got a chance at your next one.

Abbate:

Did it ever work the first time?

Haibt:

Not very often, for most people. One of the people in the group it seemed to always work for; he was a fanatic about it. One of the great stories I have is that two other people in the group, Harlan [Herrick] and Irving [Ziller], had a bet that Harlan’s program would work flawlessly—Harlan just agreed to be honest about it—and it ended up in a draw, because there was a mistyping on the key-punching. Something was spelled wrong, which the assembly program complained was an undefined symbol (it didn’t know it was a typo)—and they couldn’t agree on whether this was a bug or not! [laughs.] Harlan was saying that was a bug; he didn’t do it perfectly.

Abbate:

Had he typed it in?

Haibt:

I don’t think so. In those days, most of the key-punching was done for you by someone. You took it in and dropped it in the basket, and got it back in a couple of weeks.

Abbate:

A couple of weeks?

Haibt:

Yes. Short things in a week—you know, if you just had a few corrections.

Abbate:

So there was a room full of women doing key-punching.

Haibt:

Sort of like a typing pool.

Abbate:

And it took that long!

Haibt:

Well, at that time there weren’t many people who could key-punch, and they were hiring programmers rapidly. It was growing rapidly then.

Abbate:

You would never punch your own things?

Haibt:

Well, in fact I did eventually, because of that delay. Where our offices were, we shared the floor with another group, which did work for charities. At that time IBM did punch-card–type accounting for several of the charities; the Salvation Army is the only one I can remember, but probably the Community Chest (or whatever it was then, before the United Way) and something else. It turned out they had extra machines. They had three key-punches and three people, but they also had some verifiers and some sorters and stuff, [so when they were using these machines they weren’t using all three of the key-punches,] and I was able to borrow one of their key-punch machines for most of the time. So I sat down and punched out my own program—and came home with carpal tunnel, it would be called now! [laughs.]

Abbate:

But your program would be done earlier?

Haibt:

Yes, several weeks in that time, because it was a big chunk of stuff.

But we—the people that were just hired—were doing some practice problems and whatever. I’ve always suspected [that the reason I got put onto a real project] was because I kept going back and saying, “What can I do? What can I do? I’ve done this; what can I do next?” And they’d say, “Well, everybody’s on vacation,” and they’d try and find something. But then John Backus was looking for people, so he borrowed me and one other person for six months, to finish up this program.

Working on the FORTRAN compiler

Abbate:

And this was the FORTRAN compiler?

Haibt:

The FORTRAN compiler. I started working there in six months, and it sort of got to be a career! You know, I never went back [to my original group]. Eventually John stopped managing, after the FORTRAN project. I stayed with the successor groups and did other things, but always ended up back in compilers.

Abbate:

Now that was quite a major project, the FORTRAN compiler.

Haibt:

Yes, that was a very exciting. It’s a hard act to follow, if you start there; you know, there’s nothing like starting at the top!

Abbate:

What was that like for you?

Haibt:

Oh, it was exciting. At that time I felt very junior to the other people. I mean, it turns out that in those days, a senior person in programming was two or three years older than you were, probably—or at least two or three years more experienced; I think they were all more than three years older than me. But they were probably all 30 or under.

And by the way, if you read that article in the New York Times, don’t believe it! [ Steve Lohr, “Software: When Few Knew the Code, They Changed the Language.” The New York Times, June 13, 2001, Section H , Page 20, Column 1]

Abbate:

I haven’t read it; I’ve just seen a reference to it.

Haibt:

It’s not seriously bad, but it’s kind of sloppy—including a picture of some other group! [laughs.] Thirty men and one woman: it was not the FORTRAN group. [The Times published a correction on June 22, 2001, stating: “A picture in the special E-Business section on June 13 with an article about the development of the Fortran computer language was published in error. It showed the first programming class for the I.B.M. 704 computing system in 1956, not the I.B.M. team that developed Fortran in the late 1950’s.”]

Experience as a Woman Computer Programmer

Abbate:

How many women were in the group?

Haibt:

For the first couple years, I was the only one. So I was the only woman and much younger, but I was always treated like one of the group—except they always wanted to stand up for me! At that time our desks were arranged in a row, and each person had a little side chair for somebody to sit and talk to them. I got to be very good at walking over and then sliding into the chair quickly, because every time I walked over to answer a question, they would stand up. They would be uncomfortable sitting there with a woman standing, talking to them! [laughs.] It’s a long time ago. Things have changed—in general for the better.

Abbate:

This group was fairly small, probably?

Haibt:

Yes. It was somewhere around eight to ten. I’m trying to remember: there was

Harlan, and Peter [Sheridan], and Bob [Nelson], and Sheldon [Best] . . . I think I was the seventh. And gradually we got several more, and then after a few years, when we went to the second release, we added about five more people, and two of those were women. In fact, when I started, the department secretary was even a man. He was a temporary; he was an unemployed would-be actor. So I really was the only girl!

Abbate:

Well, at least they didn’t think you were the secretary—I hope!

Haibt:

Not then. Later on, I got strong language over the phone because I wasn’t taking shorthand fast enough when I was answering the phone, trying to take a message when I was the only one around. And I had that trouble for years. Even when I came back to work in 1972: our group moved their offices, and I went to the new office to go into my office, and some very nice gentleman from the department adjacent came over to ask who I was getting the office ready for, because he was the manager of the group in the next few cubicles. He was a little embarrassed when I said it was my office. Very apologetic.

Abbate:

So there didn’t seem to be that many women programmers? Or just not in your group?

Haibt:

There weren’t in our group. There weren’t very many. There were some, and compared to nowadays, even, there were a fair number in the training group—nothing like half, but I look back on it and I think, “You know, in those days they were hiring anybody they could find for a programmer.” It was a brand-new thing, so there was nobody trained. Computer science in colleges didn’t exist; they were just beginning to do a little in the electrical engineering department and in some math departments, which was mostly numeric analysis; it wasn’t the logical kind of programming, which is the part I like. It was “What do you do about the tenth decimal digit?” and rounding errors. Absolutely not my thing! I always said I got to be a mathematician so I didn’t have to do arithmetic; I could write “A plus B” and didn’t have to worry about the details!

Details on Programming Projects

Abbate:

Did you know anything about compilers before you started this project?

Haibt:

Nobody knew anything. They didn’t exist! Everything we did we invented on the fly. They had done a few packages that were sort of what they would call “interpreters.” John had worked on a floating-point package—that was partly how they got into the original [FORTRAN] design—and then he just hired people as they realized they would need more people. This was more than one person could do, [so] he got other people to work on that part of the thing.

Abbate:

Was there some aspect that you found particularly interesting, or that you ended up working on?

Haibt:

Well, at the time I joined the group, they had parceled out some things, and they said, “We need somebody to do some flow analysis.”

Abbate:

Which is what?

Haibt:

In order to do the optimization, you need to follow through the streams of instructions: to follow the path of execution through the program, and branches—where they combine and where they loop around, and so on. Get some figures, so that when they did the register allocation, they knew how to get six zillion quantities into and out of registers, of which you only had a few!

Abbate:

Right. So you want to determine the whole sequence of instructions . . .

Haibt:

. . . and which are the most frequent paths, because those are the ones that you want to optimize first.

Abbate:

Ah. You want to know what resources are required, in terms of memory in the computer.

Haibt:

Right, yes. And then someone else did more of the analysis of the register allocation, an analysis based on the frequencies that my part of the program generated.

Abbate:

That sounds quite complex. I mean, it sounds like you really needed to know exactly what’s going on.

Haibt:

It was challenging. It’s not drudge work. I mean, punching all those cards was no—particularly for somebody who had failed typing! [laughs.] But mostly it was exciting. I mean, not only was it new to me, but it was new to the world! In fact, after I was there for a month or two, one of the senior people around in the computing area who was not in our group said to me, “I want to ask you something. Those guys, they’re committed, so they’re not going to tell; but you’re coming from the outside. What do you think about this? Is this ever going to work?” I’ve always been sorry I couldn’t remember which of a couple people it was, because it would have been fun to [remind him]. [laughs.] Or maybe not!

Abbate:

What did you tell him?

Haibt:

I said, “There’s no question it’s going to work. We know what we’re doing and how to do it. How much it will be used, I have no way of knowing, because I don’t have any experience in the field, so I don’t know how to judge how it will be accepted. But it will certainly do what it’s supposed to.”

Abbate:

What was his concern?

Haibt:

He didn’t really believe it could be done. He thought this was a pie in the sky.

Abbate:

That literally you couldn’t have a higher-level language?

Haibt:

Well, that it at least wouldn’t work well enough so that anybody would ever want to use it.

Abbate:

It wouldn’t produce good code?

Haibt:

Oh, yes! It wouldn’t produce even usable code.

Abbate:

This is so interesting, because if you look at the history, it’s: “IBM develops FORTRAN”—and you just assume that they thought this was this great thing and they’re forward-thinking, and actually they’re biting their nails!

Haibt:

Well, there was this bunch of people with a great idea—and a lot of people who, I later learned, were not so all-seeing and all-powerful. [laughs.]

Abbate:

How soon did it become apparent to them that it was going to work?

Haibt:

It was certainly apparent to us before we finished. As soon as we got code that was even slightly usable, we had some people from the National Weather Service who were doing computer stuff who came and said, “We’d love to be your guinea pigs. If you will let us, we will take your output, even though you haven’t got the register allocations working.” This was at a stage where our program just came out with symbolic registers from one to a zillion, you know, and hadn’t yet squeezed them into the three registers on the machine; and they essentially hand-allocated them, and jiggered up the assembly program.

Abbate:

So they were so desperate for [a higher-level language] . . .

Haibt:

They were. And even with that much hand-work, they felt it saved them an enormous amount of time. Then when we came out with the original public paper (although it was known [before that]), they worked very hard to get some people to do some test cases to present at the National Joint Computer Conference—some of the aircraft companies and stuff, which were the big computer users in those days. [The companies] got two sets of people to write some of their programs: one in assembly language and one in FORTRAN. They only got a few, because it was one of those, “Are we going to get it actually running before then?” kind of dates. But even then, at that very early stage (which would now be considered end of alpha test, beginning of beta test kind of stage), they found that the FORTRAN programs actually ran slightly faster—which was very important in those days; none of this gigabytes of memory and megahertz and whatever—and that it was a fraction of the effort to do. So it was faster to write and ran better.

Abbate:

So the users were ready to adopt FORTRAN.

Haibt:

Yes. The people who actually tried it were convinced. There were a lot of people who would look at the code and say, “I can do better by hand; they can’t possibly do it as well.” And the answer is: Anything the compiler could do, of course you could do by hand. It generates a list of instructions; you can write the same instructions by hand. The thing is, it can do a lot more analysis, and analyze a much bigger span, than you can do in your head—and it can update it when you change something. When you do it by hand, you try to optimize the important parts, but you’d never finish everything if you tried to do every little bit of it perfectly, in the sense of “fast as possible.” It’s one thing to do little test things in class: “There are ten instructions. See if you can do it in nine.” It’s a lot different when you’re talking about tens of thousands [of instructions], in those days, for a big program. But when you had programs that ran for hours, a small percentage improvement was important, and a big percentage was very important. The difference between three seconds and two seconds, or even three minutes and two minutes, isn’t nearly as important as between three hours and two hours—particularly at six hundred dollars an hour!

Abbate:

Right. [laughs.]

Haibt:

Now you can buy your own PC to do that.

Abbate:

What happened when the group finished that?

Haibt:

Well, eventually we did a second release, and then sort of a third version of it, which was internal only, which had some character instructions and stuff, because originally it was very much a mathematical language. It still is, but much less so.

Then we were sort of in the position, looking back on it, of being a good group looking for a project—because we didn’t want to split up, because we all liked each other and thought highly of each other. It was a marvelous group to work with!

Some of [us] went off into doing the logic program. There was a young man at Columbia who solved one of [Emil L.] Post’s conjectures; I think he did what was called the “two bit problem,” which was a program learning thing. You generated a random sequence of bits that was for a two-bit simulated computer, and you made random changes in your [sequence], trying to score the result—to get a workable program with the right answer—by trying to decide which changes were good and which were bad. So at one time they were going to try and do a sixteen-bit version. Originally [they were going to do this] with one other person, who left the company shortly after we started, who would do an automatic flow-charting program: take a program and generate a flow-chart, because at that time that was the way one documented a program. Now you don’t hear so much about flow charts, but that was what they had then. So [when that person left,] I did that.

Abbate:

The flow chart?

Haibt:

Yes. It almost eventually got out the door of the company, but it was killed at the last minute, after they had the publications printed and everything. But by then I wasn’t any longer associated with it; it was gone to the Development Division. I had worked with them to do it, but they eventually made the call.

Then after that I got involved in a virtual memory project, which was the origins of virtual memory.

Abbate:

So this was another very new thing at the time?

Haibt:

That was another exciting, nobody-had-done-it, breakthrough-type project.

Abbate:

And that was . . .

Haibt:

That was in the early ‘60s. It was based on a modified IBM 7044, and it preceded the 360 series, which was announced while we were working on this, but even within the company it was fairly secret for a long time. And we got that working, and it later became the basis of the 360 Model 67, which was the first to have a virtual memory. It then became CP/CMS 67, which became VM/CMS. So that was exciting.

Then I had my daughter, and worked part-time until she started first grade.

Balancing Family Life and Work

Abbate:

So you’d gotten married at some point.

Haibt:

Oh, I got married early on. Yes, I got married about a year after I got out of college. I had been married for about ten years at that point.

Abbate:

Was it difficult balancing work and family? Did you take time off?

Haibt:

Yes, I had some problems. Not serious. My daughter was extremely healthy, but she was about four pounds; she was early. And I was not ready to come back to work at—in that day, three months was the maximum leave you could have; three months after the baby was born. They also kicked you out of the door three months before the due date—almost literally! They sent a note to my manager, which I saw, saying I could not work beyond Tuesday, January 11th.

Abbate:

So you had to tell them the due date?

Haibt:

Yes. And I didn’t know enough to fib in those days. I was too honest; and they required notes from the doctor.

Abbate:

Of course, in your case, the due date was sooner than you thought.

Haibt:

Yes. And I couldn’t work past 5:12, which was quitting time. Under no circumstances could we extend it, and under no circumstances could I work late that last day! [laughs.]

Abbate:

Oh, just the last day.

Haibt:

Yes. So I went from working probably 60 hours a week to nothing, overnight. Because we were getting the project—it was just about running at that point.

Abbate:

This was the virtual memory?

Haibt:

Yes, the virtual memory thing. At that point I was literally working as much as I could. I’d get up in the morning and come to work, and work until I couldn’t work any longer; go home, fall into bed; repeat! [laughs.] And I was usually in at least one day on weekends for a while. Fortunately my husband was also working here.

Abbate:

So you got to see him.

Haibt:

So we had lunch together! [both laugh.] Lunch and supper, because he was also working crazy hours. He was working on a project which required taking over the machines, so they had to schedule their machine time from three in the afternoon until midnight, or something. He usually came in pretty promptly in the morning anyway, so we were both working weird hours.

Then the manager of the group at the time observed that since we had a terminal-oriented system, maybe I could put a terminal at home—which was long before any such things ever happened. So I got a terminal at home and worked part-time, after Carolyn was born.

Abbate:

So this was after the three months of maternity leave?

Haibt:

Yes. At three months I was in no way ready to come back to work—just not up to that.

Abbate:

But you were able to work part-time. Was that unusual?

Haibt:

Yes, it was. In fact, when they proposed this and it got up to the Director of Research, who was a Corporate Vice President at that time, reporting directly to the Chairman of the Board: When he decided this was a good idea, he’d like to make it happen, it still took five months or something to get the approvals through [IBM headquarters at] Armonk.

Abbate:

You were the first person they had ever tried to do this with?

Haibt:

Yes, right. There were some bureaucrat types who were afraid that I might fall down the cellar steps and sue IBM for workman’s compensation, or who knows what.

Abbate:

So it wasn’t the part-time part so much as the work-at-home part?

Haibt:

It was the combination, I think; but mostly the work at home.

Abbate:

Interesting. So, for those five months you couldn’t work?

Haibt:

Well, I didn’t really want to anyway. A small baby takes a lot of attention. But after that, it was nice. I had a terminal at home, so when I put Carolyn down for a nap, I could go work for a couple of hours. Or I could go start the laundry—which of course there’s a lot of, with a baby!—just go throw things in the washing machine and come and work for a few hours. So that was very nice. I managed to do that for several years: three, I think.

Then there was a crunch, and there wasn’t really anything suitable. But Lou happened to have lunch with somebody who had come to IBM from Computer Usage . . .

Abbate:

Who was Lou?

Haibt:

I’m sorry; Lou was my husband, who died just last year.

Abbate:

Oh! I’m sorry to hear that.

Haibt:

But it was very nice: we were working on the same things—and mostly, for the first half of our careers, in the same building, so that was very handy. Then he moved to an office in White Plains, but we were still doing similar things.

Anyway, he was talking to somebody who’d been working for Computer Usage Corporation, which was big at that time—a consulting, job-shop-programming–type place—and somehow some mention of it came up, and he said, “Oh, I bet the guys I used to work with could use her.” So I worked there for a [while], also part-time. Not at home—well, paper-and-pencil at home, not a home terminal. I would write my programming at home, and then bring it in to get it punched and run it and stuff, while Carolyn was in nursery school.

Abbate:

You’d been laid off from IBM, more or less, for lack of work?

Haibt:

Well, it was contract-by-contract; and after three years there was nobody who had money in a project that could hire me. I mean, they were all limited things.

Abbate:

So all your work at IBM was for a fixed-term contract, and you had to sort of hunt around for a new one?

Haibt:

Once I went to part-time, they didn’t have such things as part-time regular employees. Now they have all sorts of different things, but at that time you were either a full-time employee or you were on contract with a limited time.

Abbate:

Okay.

Haibt:

And so they just ran out of things that they could do and funding for doing them.

Then I worked for a little startup company for a while.

Abbate:

This was after Computer Usage.

Haibt:

After Computer Usage, yes. About a year after I started with them, they had overexpanded and sort of imploded; closed their Mount Kisko office.

Abbate:

And so then you went to this startup.

Coming Back to IBM Full Time

Haibt:

So about that time, I went to this little startup that was some former IBMers who had a notion of doing medical records, medical billing, and selling it to doctors. A time-sharing sort of service, of which there are lots now, but they were fairly unique at that point. Then they fell apart, after about a year or so, because they had—I mean, it was a sort of an early version of the dot-coms; nowhere near as crazy, but very small-scale. They were a little over-ambitious, and didn’t have a good sight on what they needed to do to get going and get their first customers and so on.

Then my husband got a sabbatical. He was teaching up at SUNY Binghamton, starting up what they called a “School of Advanced Technology,” a sort of computer-related graduate school. I had worked for another contract for IBM after the startup, and so I finished that up. We were up there [in Binghamton], and it was a new place, and Carolyn was starting school and stuff, so for the first year we were up there, I did not work. The second year we were up there, not only was I interested, but Lou was pushing me, because he could see—you know, I would go to the ACM meetings (the local chapter) and the various computer-related things. So he was saying, “Why don’t you do something?” Kind of pushing me to talk to the university, which had very little to offer as far as programming jobs, but I interviewed with IBM.

Abbate:

IBM had a Binghamton branch?

Haibt:

Yes, what they called at that time Federal Systems. Endicott was where it started, so they had a big branch there. And I applied for part-time work, but nobody from personnel even bothered to talk to you. That was when Ph.D.s in physics were driving taxis and tending bar!

Abbate:

This was in the ‘70s ?

Haibt:

This was seventy-one, I think, which was another one of these dips. Anyway, I went to talk to the people in Owego, and the personnel guy was so nice and so glad to see me, because he said, “We’ve now got four slots.” Two of them were for RF (which I presume is “Radio Frequency”) engineers, and two were for programmer/architect/whatever. But he was so glad to have any slots that he could hire anybody for; he hadn’t had any for a couple years! [laughs.] It’s hard on somebody in personnel who would like to hire people.

Of course, I’d never really looked for a job in my life, so I didn’t quite know what to expect. But I was taken upstairs to meet the fourth-line manager, which I now know is typical; if you’re a good candidate, then you go see the high-level guy, and then they can observe that you don’t have three heads or something, I think! And he said, “You know, you describe what you’ve done, and this isn’t quite my field, but there’s a guy who used to work for me, and this is right in his alley. So if you wouldn’t mind, I’d like to have you talk to him, just because he could better understand what your experience was.” So I went to talk to him, and he was kind of a clown who liked to make a big show of things, and he was saying, “There’s somebody I’d like you to talk to, but I really shouldn’t do it because I haven’t got a slot, but you’d really like him”—going on like this. Well, eventually I did talk to the man, who became my manager then: Tom [Philbin?], who it turned out had just come to Owego to work on compilers the week before, after having spent several years here in Research. I didn’t know him; he didn’t know me personally; but he know of me, because he was working with all the people that I had worked with for all these years, and he knew Lou. He had been working very closely in the group I had left, essentially. So he’d heard about me and knew all about me! [laughs.]

Well, this was on Tuesday, I think, and on Wednesday the personnel guy called me to say, “I don’t have anything to say; it’ll be a while, but I just wanted you know I haven’t forgotten you, and we’ll get back to you.” And Thursday I got an offer—to start Monday! [laughs.] And they really wanted me full-time, not part-time; and they really wanted me this Monday, not next Monday; and because they understood I had child arrangements to make, they’d understand the first day if I came in a little late. So that was sort of interesting—because by then I’d been at home for six years or something, and I had one dressy dress, so I had one day to get a working wardrobe! Because I had been wearing pants, on account of when you have a little child, you’re bending over all the time; so my wardrobe was long on slacks and very short on work-type dresses.

Abbate:

Did you go back full-time?

Haibt:

Yes. I ended up going back full-time, and worked there for a year. It was even more flattering [to be hired] because they knew I was only there for a year, because Lou was on sabbatical. In fact, I later found out that somebody in the bureaucracy—not the guy I was working for, whom I know wouldn’t be one to do something like that—had called Research to say that it was clear I would be moving back down here and wanting to transfer back to Research, and would Research be willing to take me back? And fortunately my friend said, “We’d love it!” As I said, there was a time limit [for contract work], there’s no [permanent part-time] jobs, it’s the way things go.

Abbate:

What project were they so eager to have you for?

Haibt:

That was a compiler, another FORTRAN compiler, for an Air Force computer that had some romantic name like “ANQ-47” or something! It was used in the AWACS system: Airborne Warning And Control System. It was the radar stations, and the planes, and whatever.

Abbate:

Was that related to the SAGE system?

Haibt:

It was later than the SAGE. SAGE was late ‘50s, early ‘60s, somewhere in there, because that was originally a 704-based thing. But it was that sort of stuff.

Abbate:

So you did another FORTRAN . . .

Haibt:

Yet another FORTRAN compiler. I had done [another] one—I had managed one; I hadn’t done it myself—as part of the virtual memory project, because we did a system for the machine.

Then, when I came back here, there was division called ASD, which was working on a project that was then called “Future Systems.” This was a little too ambitious and never actually came out as a product, although pieces of it came out in various products; essentially they stayed with the 360 family and improved it to the 370 and the 390 and whatever it is now.

Then I came back to Research and worked on a study and paper about large-scale program development. “What are the problems? What ought to be done to make it easier to write humongous programs?” It really is different than writing a small program.

Abbate:

How did you work on that? Were you going around asking people?

Haibt:

Yes. There were four of us, and we went around and interviewed people in all the development labs around the country—and a few places outside of the country, although I never got sent on those trips; but mostly it was in this country—talking both to the management and the programmers to see what their problems were. That was very well received, but made very little impression, unfortunately.

Abbate:

It sounds like it would have been interesting for you to see what everyone was doing.

Haibt:

Yes. And we had some opinions about how things should be done, and made some proposals for developing things in a test situation, where you could compare algorithms and ways of doing things and so forth. But also, the way they were setting up large programming projects was a very linear sort of thing, like you would do a manufacturing line. [The managers’ notion was that] it was something you knew how to do, and you just had to do more of them: you build this part and then you build that part; you could schedule everything and you know how long it’s going to take. But producing a program is not like running an assembly line. In the analogy to the assembly line, there’s the place that duplicates the tapes or disks, or prints the manuals; but developing [software] is really an invention process, to a large extent. It’s very hard to predict, and it’s an iterative thing inherently, was our feeling.

Abbate:

How did you think it should be done?

Haibt:

Well, much more realistically. They had charts and processes and whatever, and what it boiled down to was, when some date [was reached], whatever they had was (as we said) “thrown over the wall into Test.” It was done because the calendar said it was going to be done on this date! [laughs.] Some places were extreme: one group wrote the code down on paper, and handed it through a window to somebody who punched in the cards (at that time they were still using cards), and then some other group tested it—and there was no feedback! So if you never find out what your mistakes are....

Abbate:

Was that their attempt at rationalization, to have this division of labor?

Haibt:

Yes, because it got up to a certain level, and [the managers thought that] programmers were all interchangeable. Big projects would move around, for load balancing or whatever. They would move from one lab to another, or from California to Europe or something, and the people wouldn’t go with them: just some new set of people had their hats with new names painted on them. And it just didn’t work that way! All the knowledge about how this worked, and how it fit together, was in people’s heads. It couldn’t be documented. (It certainly wasn’t; it’s not clear how much of it could be documented.) The idea that programmers are not equivalent to a typing pool, for example, or an assembly line, was not a factor in any management decisions.

It’s an art, not a science—and sometimes a black art! [laughs.] It’s not cut and dried, and you can’t just drop a big project in the lap of a whole new set of people, because it will take a year for them to figure out what’s going on. About that time, it would be moved again.

Abbate:

But they didn’t follow your recommendations?

Haibt:

Not too much. Some of them, but we didn’t have the impact we had hoped.

Abbate:

Did you find there was a change over time in how programming projects were managed?

Haibt:

Oh, very much. When we started, everything was small. Our project was considered a huge project.

Abbate:

With eight or ten people.

Haibt:

Yes. [I worked for] several years with programs that tended to be small, individual things. And now there are casts of hundreds, if not thousands!

Abbate:

How are they managed now?

Haibt:

Better than they used to be! [laughs.] My daughter is also an IBMer. We laugh and say that she didn’t know there was any other way to live. She had a terminal at home before she could walk!

Abbate:

Wow!

Haibt:

You know, both parents doing it; and hearing about it, because we’d talk shop a lot. All programmers do—which is hard on non-IBM spouses! [laughs.]

She’s doing very well. She’s more in the Development Lab, whereas I was always in Research. She would have loved to be in Research, but they had the two-job problem.

Abbate:

She and her husband?

Haibt:

Her husband, yes.

Reflecting on Women in Computer Programming

Abbate:

Interesting.

Do you think it’s gotten easier for women to get into programming since you started? Or harder?

Haibt:

Well, to get into programming, I think it was easy back then because there was nothing established. They’d take anybody they could get whom they thought was qualified—or even if they hoped to be! They took bridge players, chess players, geologists, philosophers—you name it. Then there was a period when I think it was much harder. But now I think in a lot of places, at least, they’re promoting women much more, at least at the lower level. There are a few very visible high-level [women], but it’s still tremendously unbalanced—even in Research, where there’s now, I think, a second woman Vice President. There was one a couple years ago who moved on to one of the divisions, and now we again have one. There are still very few women in research management.

Abbate:

Did you feel that as a woman it was hard to get promoted or to move past a certain level?

Haibt:

Well, it’s a little hard for me to say. I never thought about it. I think partly it was the kind of person I was. I did become a manager just before I left and had the baby, but there were very few women in management then, and for a long time afterward. I don’t think it was so much explicit as just [that] we didn’t think about it; and certainly nobody ever encouraged me to speak up or do things. How much of that was because I was a woman, and how much of that was because I was such a quiet, unaggressive person, I’ll never know.

Abbate:

Did you have any mentors or role models?

Haibt:

Not really. Not in those days.

Abbate:

Well, maybe not formally, but . . .

Haibt:

Well, I had people I worked with and enjoyed working with, and worked with for many, many years. I sort of gradually got into things.

Working on Other Programming Projects

Abbate:

Let’s see . . . You described how you did this report on programming. What happened after that?

Haibt:

Then I was working on a program analysis project, in which we were trying to understand programs and convert assembly programs into compiler programs.

Abbate:

You mean, take existing programs and automatically . . .

Haibt:

. . .translate them—or at least ninety percent automatically translate them—into high-level programs; which would then, presumably, be easier to clean up and maintain. That ran into problems because of resources. It got to the point where I thought it was going pretty well, but it needed more than one person to do it. I also had gotten interested in another project, which was a system programming language, which really had some PETRInet underpinnings.

Abbate:

PETRInet ?

Haibt:

PETRInet. “Petri” is the dishes in chemistry.

Abbate:

I haven’t heard of this one.

Haibt:

That’s a logic sort of thing, good for parallelism, and a way of keeping track of when all the inputs to something have been made available, so you can process them and generate something. Well, that kind of petered out, partly because Bob Nelson, who was the head of it, died. About that time I’d also gotten involved in yet another FORTRAN compiler with the 801 system, which is the origin of the RISC architecture: so yet again, a breakthrough project.

Abbate:

RISC is “Reduced Instruction Set Computer.”

Haibt:

Yes, Reduced Instruction Set Computer. They wanted a FORTRAN compiler, and so I got involved in that, and that became a joint effort with the Toronto lab, which was the development lab. So [the project involved] the research ideas and then the development and production.

Haibt:

How different was it, doing a compiler for RISC?

Abbate:

Not different at all! I mean, there were some better techniques, after thirty years of research—because there’s now a whole science of computing. Now they teach it as an undergraduate course; they don’t have to invent it! Everything we did, we invented on the fly.

Abbate:

Did you ever do any teaching?

Haibt:

No. I’ve done some one-on-one, and my husband and I and another researcher ran an Explorer Scout troop for a number of years: high school kids. So it was sort of hands-on teaching. Mostly we sat there while they were on machines, doing things. Some of it was teaching—more like tutoring, I guess. I would not like to teach a class!

Abbate:

Even on compilers?

Haibt:

Even on compilers. I could talk to somebody, but I once it becomes presenting, I fall apart.

Abbate:

Well, most people don’t relish that.

Was that your last FORTRAN compiler?

Haibt:

Yes. Well, I was working on that, and then that sort of finished up—became a development project. I did a little bit on some other things. Some Java, a performance study, and little bits and pieces of other compiler-related things: some vector stuff for parallel machines and things.

Abbate:

Is that quite different?

Haibt:

Not really. More of the same.

Abbate:

So once you really know compilers, it’s just [more of the same].

Haibt:

Yes. I mean, they’re all interrelated. Some of the stuff that we were doing on these fancy complexes of a zillion machines are different in scale, but in fact when you look at what’s really being done, it’s the same sort of thing we were doing back in the early ‘60s on what we then called “automatic storage allocation”—except then it was core memory and tapes, and now it is several levels of cache, and real memory, and disk, and networks. So the scale is different. We’re talking about microseconds or picoseconds instead of seconds, but it’s still the same idea: if your CPU is faster than your memory, then you want to try and arrange things so you get the right pieces at the right time—fetch them ahead from wherever they’re stored in mass storage into a faster, local storage. So in some ways it’s very different, and in some ways, the basic outline of the problem is the same.

Reflections on Status of the Field of Computing

Abbate:

That’s interesting.

How do you think the field has changed over time?

Haibt:

Well, of course one thing is that you used to know everybody and everything—at least superficially—and now, I don’t even know the people on this floor! [laughs.] And certainly nobody knows all about computing. You don’t even know all about compilers. You’re lucky if you know all about some little piece of it, because it’s just so much bigger. It’s like people who drove cars used to know everything that was in them, and fixed them—had to fix them; several times on a Sunday drive, probably! And now, you get in a car with automatic shift and automatic instruments, and hope the light goes on to tell you if something goes wrong . . . which is both good and bad.

Abbate:

Right.

Haibt:

It’s good that you don’t have to know all that stuff; you just get in.

In the computer world, they talk about appliances.

Abbate:

Right—although that hasn’t really caught on, I think.

Haibt:

Well, compared to what it used to be, it certainly is. I mean, when you sit down at a P.C. now and use it, you don’t know what’s going on under the surface. You have icons, and it does things, but you’re certainly nowhere near the bits-and-bytes level.

Abbate:

But it’s not quite like your refrigerator yet.

Haibt:

No. But . . .

Abbate:

Have you been active in professional societies?

Haibt:

No.

Abbate:

You had mentioned going to ACM meetings . . .

Haibt:

Well, when we were in Binghamton, I would go with Lou to the local ACM chapter, sort of keeping my hand in. It was small. I think early on I went, for various reasons, to one or two [meetings]. I don’t like going to talks, and I was never a good committee person. But that’s another instance of having no notions of career advancement. I didn’t have anybody who said, “If you want to get ahead, you should make yourself known,” and so forth.

Abbate:

Did you publish things under your own name?

Haibt:

A few things. Only one or two outside things; some inside things. Mostly I just did.

Abbate:

Got things done.

Haibt:

Yes. I don’t like to write, any more than I like to present.

Looking Back on Career in Computing

Abbate:

What do you find the most satisfying aspects of working with computers?

Haibt:

The thing I think of, looking back, that I’ve [enjoyed most] was doing things that people didn’t necessarily think could be done—and working with some nice people. The big projects that I mentioned have all had that: they were breakthroughs in the computing world, and they were done with people I enjoyed working with. They were exciting to work on, as well as to see the results. You know: “Hey, we can do this! This is going to make things better.”

Abbate:

And did you get to see people actually using them? A few years later, thousands of people are working on something you wrote?

Haibt:

Right. There’s a little something-or-other when your daughter goes to college and takes a course in FORTRAN, and has this instructor talk about having met those crazy guys and whatever! And she sits there and smiles to herself, and says, “That isn’t the way I heard it!” She didn’t say anything out loud, but sooner or later people did find out. Somebody saw some names or something and made the connection—because Haibt is not a very common name.

Abbate:

So that’s your married name?

Haibt:

Yes. [My maiden name was Lois Mitchell.] Back then it was occasionally done [to keep one’s maiden name], but very seldom. And there was also a Libby Mitchell in our area. She was actually Grace Elizabeth, but of course everybody thought “L. Mitchell” was Libby, because she’d been around a few years more than I had; and she later in fact joined the FORTRAN group. But I didn’t really think very seriously about keeping my name—although sometimes I’ve thought it would have been easier to spell!

Abbate:

So there would have been two L. Mitchells.

Haibt:

But there certainly were not two Haibts. And my husband was Luther, known as “Lou.” So it took a while for us both to get in the phone book.

Abbate:

Because there were two L. Haibts.

Haibt:

Nobody could believe there were two L. Haibts in the same department—same department number!

Abbate:

Maybe they should have used two initials.

Haibt:

Right, well, they’re “M” and “H,” which tend to look the same anyway. And for years, every time one of us had some reason to change the phone book, the other would disappear, because somebody would proofread it and say, “Oh, they didn’t take out the old entry!” The first time we both made it in the phone book was back at the Lamb Estate, which was a temporary location while they were building this building. We all knew everybody there, because there were only forty people or something in that building, and I said . . .

[DISC 2]

Lou was in Poughkeepsie, so we didn’t have, at that time, the problem of being confused with each other. It wasn’t till the combined research here that we began having problems.

Abbate:

I guess you can have two John Smiths or something, but not two L. Haibts.

Haibt:

Right. Well, right now the group that I’m attached to has two K. O’Brians. There’s a Kevin O’Brian and a Katherine O’Brian, who are husband and wife. They’re going through the same thing. But at least O’Brian, people could imagine that there’s more than one of; but they couldn’t imagine there was more than one Haibt. They couldn’t really quite cope with one! [laughs.]

Abbate:

Well, hopefully nowadays they’re a little more used to having married women there.

So, let me think: Have we missed any career highlights?

Haibt:

I don’t think so. I believe we went through in chronological order.

Retirement

Abbate:

What are you doing now? Are you still consulting?

Haibt:

Well, I was I having some problems; I was just getting extremely tired, which meant it was hard to get interested and successfully do any work, so I decided it was time to retire. I got the Emeritus appointment last summer, a year ago, and I was coming in about once a week to read my mail and check up on the newsgroups and stuff, and about to get back in doing some puttering around with some useful things: writing up how to write a compiler front-end, which was one of the things that I got very good at. Other people don’t necessarily know how to do it right.

Abbate:

That’s a good question.

Haibt:

Not great inventions, but ways of proceeding, tricks of the trade, and so on.

Abbate:

Because you have all this experience that’s sort of tacit knowledge.

Haibt:

That’s right. I know some ways to set things up to make it easier for yourself as you’re doing the front-end.

But then last December, right after Thanksgiving, my husband was killed in an accident.

Abbate:

Oh, goodness.

Haibt:

And that turned my life upside down. So I’ve had my hands full since then: all the paperwork, and all the lawyers. And so I haven’t—it didn’t quite go the way I planned.

Abbate:

Compilers were not the first thing on your mind.

Haibt:

The story of my life is one of those little signs that says, “Life is what happens while you’re making other plans!”

So coming to work has . . . I have home access; I have my Thinkpad at home; so I have been logging on regularly all this time, but I haven’t gotten into the office very much. And I keep saying, “I’m going to start; I’ll get in next week.” And then next week doesn’t happen, because somebody calls up or wants copies of papers I haven’t seen since 1960-something.

Abbate:

Does your daughter live nearby?

Haibt:

She lives down in Raleigh. She was able to bring her computer and access and spend a month with me to help me get it straightened out, and she’s been up about once a month—because she has business [here]; being at IBM, she has business in Armonk or customers in New York. So she will often stay at, as she says, “Chez Mom” for her hotel, or at least spend the weekend while she’s here.

Reflections on Working for IBM

Abbate:

Do you think, overall, IBM has been a good place to work?

Haibt:

Yes. It’s not like it was, though.

Abbate:

In which ways?

Haibt:

It’s much more hard. “What have you done for me lately?” It used to be, “If you work hard, we’ll do our darndest to find you a place; if you can’t do the job, we’ll find you a job that you can do, as long as you’re trying.” And now it’s: “We’ll lay off another couple of tens of thousands of people this week. Yeah, we’ll change your retirement plan. Yeah, we’ll cut the medical plan.”

Abbate:

Times have certainly changed.

Haibt:

Yes. I think it’s probably still better than most anywhere else—although I haven’t got really much in the way of comparison. And I don’t get the same feeling.

Abbate:

Well, it’s certainly changed.

Haibt:

You know, it’s “Watch your back, and make sure you have a good resume. Look for another, better job.” Whereas it used to be everybody was happy to get to IBM and thrilled to stay there—because it was the best and most exciting place.

Abbate:

Do you have any advice for young women considering going into computing today?

Haibt:

Go to it! Have fun! It’s a standard question. I haven’t got a good answer, in spite of the fact that my daughter is doing it. But she’s certainly under a lot more stress and strain than I ever was.

Abbate:

So you think it’s harder for her?

Haibt:

Partly the kind of job she’s got. But there’s much more pressure to produce and work long hours. We worked tremendously long hours, but it wasn’t because there was any feeling that we had to, that somebody was expecting us to.

Abbate:

You just wanted to?

Haibt:

It was because it was so much fun, and you wanted to see it work! It’s not that much different than doing a jigsaw puzzle, when you’re near the end and you’re going to stay up to finish it; or reading a book, that you want to see how it comes out. It was much more that feeling, rather than that somebody was going to evaluate you on whether you made some deadline. And there certainly wasn’t the “marking on a curve” kind of performance evaluations.

Abbate:

Well, it sounds like you’ve had a lot of fun.

Haibt:

Oh, yes! Until the very end I kept working. I wanted to. The first time I became eligible for one of these early retirement incentives, [I thought], “Why would I want to do that for? What would I like to do? I’d like to play with the computer, and do things well; right now I’m getting paid very well for doing it, and they update my equipment regularly, instead of buying my own. I have fun things to do that are going to be useful to somebody. What do I want to retire for?” [laughs.] And I’ve got nice people that I work with every day and I enjoy seeing. What would you retire for? To do what you want, spend your time with people you enjoy? Why retire? But eventually the time came when it was just physically too much. So we’ll see how that plays out.

What else? What did I not answer?

Abbate:

I think you’ve probably answered all my questions, unless there’s something I’ve missed.

Haibt:

I don’t think of anything else. If you have other things that you come up with, I’ll be happy to talk with you again, in person or on the phone.

Abbate:

If one wanted to encourage more women to get into computing, do you have any idea what the best strategy would be?

Haibt:

I think probably having women visible. Certainly I didn’t have any role models or anything. I got into it knowing I was unusual, to put it mildly, because I was interested in and very good at math. So I just did my thing, and fortunately, early on it was not a problem. I was one of the group, and treated like one of the group, not like somebody special or strange—particularly not strange.

Later on there were occasions when I felt certain individuals were not treating me appropriately, but never a serious problem.

Abbate:

Not real overt sexism . . . ?

Haibt:

Not overt, and not people who had any particular power related to me. I always had enough people around who knew me and knew what I could do that I didn’t feel I had to prove a lot.

Abbate:

You felt you were judged on your work?

Haibt:

Yes, pretty much. But I think, looking back, as far as how to be more visible, or gain more status in some sense: probably if I had been a male, I would have had more coaching. But I don’t know. It’s very hard to tell how much of that was personality and how much [was gender]. I’m sure they were both involved quite a bit.

Abbate:

Maybe if you were a man you would have been pushed into it even if you didn’t want to do it.

Haibt:

Yes; or I think you would have at least gotten suggestions as to: “If you want to get ahead, do this,” or, “You should make more of an effort to write papers or give talks,” or whatever.

Abbate:

Was there any aspect of the corporate culture that . . . You know, they must have had picnics or whatever; did they have anything that seemed inappropriate for women?

Haibt:

No, not that I was exposed to.

Abbate:

It wasn’t like they had a big IBM football game that you had to play or something?

Haibt:

No. Fortunately, I was never in a group that did that kind of thing. There probably would have been more of it in the Sales atmosphere. Research was always kind of an oddball place in that respect. It was not much in gung-ho types.

Abbate:

Were there any women in Sales?

Haibt:

I think there were some. Very few, back then. They mostly were in a separate group called System Service Reps, which were the ladies in white gloves and hats who went around and taught the secretaries how to use electric typewriters, and that kind of thing.

Abbate:

Right. I’ve heard about that.

Haibt:

And then the men had a group called Applied Science Reps. They helped the sales force; they were more scientific than the usual salesmen. In fact, most of the customers then were very scientific. So these were people who could talk to the customers technically and help them with that part of the problem, and while they were at it, helped sell IBM machines.

Abbate:

So the Research Division was probably the most gender-integrated part of the company, maybe?

Haibt:

For a long time. Although—partly because of our location—there were not many women, and it was a long time before there were any women managers. Then there were very few; and then at one point they made a real push and got a number of them; and then that kind of faded, and all of a sudden there weren’t any anymore. Well, I shouldn’t say “any,” but almost none.

Abbate:

Was that in the ‘70s? Was it an Equal Opportunity impetus to get women in management? Or was that later?

Haibt:

I don’t recall. It was probably in the ‘70s that it was fading, because nobody was pushing it.

Abbate:

So they were pushing it earlier.

Haibt:

When was it? Early ‘70s? ‘60s? Late ‘60s, maybe? At some point there was a push to have women managers, I think. They made a number of women first-line.

I was—not as part of that push, but I was made a manager, just before I left; which was very nice. Even though they knew I was pregnant and leaving shortly, they still made me a manager for a couple months—and it was sufficiently unpleasant that I said, “Never again!” I had two people on probation—one more or less formally and one informally; real problems—and one alcoholic who didn’t like working for a woman. It was in a situation where I sort of had all the responsibilities but none of the authority. So I just said, “That’s not my thing.”

Lately they’ve done another push—I think partly because Lou Gerstner was very strong on it. Apparently one of the first things he did, when he came to the company and became acquainted with the Fellows Program, was he looked around and said, “How come there are no women Fellows?”

Abbate:

Really! So that was his idea?

Haibt:

No, wait: I think Fran [Allen] had just been made a Fellow.

Abbate:

I know she was the first.

Haibt:

She was the first. And I think it was, “How come there’s no one else and they took this long to make Fran a Fellow?” And now there are a number of them, and regularly appointed. So I think it was not deliberate or conscious. It was just . . .

Abbate:

They promoted people like them.

Haibt:

Yes, that kind of thing. And a lot of the managers just didn’t see. They were just completely blind—and would have been upset had they been accused of it, no doubt.

Abbate:

All right. Probably we can stop there.

Haibt:

Okay.

Abbate:

Thanks so much for talking with me!

Haibt:

Thank you for listening to me ramble! [laughs]

[END OF RECORDING]