IEEE
You are not logged in, please sign in to edit > Log in / create account  

Oral-History:Emily Willbanks

From GHN

(Difference between revisions)
Jump to: navigation, search
(Mentors and General Thoughts about Working with Computers)
(Family / College)
Line 27: Line 27:
 
INTERVIEWEE: Emily Willbanks <br>INTERVIEWER: Janet Abbate <br>DATE: April 5th, 2002 <br> PLACE: Emily Willbank’s home in Los Alamos
 
INTERVIEWEE: Emily Willbanks <br>INTERVIEWER: Janet Abbate <br>DATE: April 5th, 2002 <br> PLACE: Emily Willbank’s home in Los Alamos
  
===Family / College===
+
===Family and College Education===
  
 
'''Abbate:'''
 
'''Abbate:'''

Revision as of 15:50, 25 October 2012

Contents

Emily Willbanks

Emily Willbanks was a scientist at Los Alamos from 1954 to 1990. She graduated from Duke University in 1952 with a Bachelors Degree in Physics and Math and earned a Masters Degree in Physics from University of New Mexico in 1957. She passed away on February 8th, 2007.

Her first job after graduating college was working for Pratt and Whitney Aircraft. While there she participated in a feasibility study about nuclear powered aircraft. She left Pratt and Whitney after two years to work at Los Alamos. Her first job there was to help in the design of different weapons systems. In time, she moved to Computer Division and spent the majority of her time designing and maintaining the storage systems at the laboratory. She had a brief interlude where she worked with Weather Control in England set-up their own storage systems.

The primary focus of this interview is her time at Los Alamos. Willbanks goes into detail about the computers they used and how they designed their own computer codes at the lab. Additionally, she discusses her relationship with the other scientists and how the different divisions worked together.

About the Interview

EMILY WILLBANKS: An interview conducted by Janet Abbate for the IEEE History Center, April 5, 2002.

Interview #635 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, 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:

Emily Willbanks: an oral history conducted in 2002 by Janet Abbate, IEEE History Center, New Brunswick, NJ, USA.

Interview

INTERVIEWEE: Emily Willbanks
INTERVIEWER: Janet Abbate
DATE: April 5th, 2002
PLACE: Emily Willbank’s home in Los Alamos

Family and College Education

Abbate:

To begin at the beginning, can you tell me when you were born and where you grew up?

Willbanks:

I was born in November 1930, and I grew up in Fort Lauderdale, Florida.

Abbate:

What did your parents do for a living?

Willbanks:

My father was Superintendent of a private estate on the beach, and my mother just was a housewife.

Abbate:

Did you have brothers and sisters?

Willbanks:

No, just me.

Abbate:

What was it like in Fort Lauderdale in the thirties?

Willbanks:

Well, it was pretty peaceful. It was a good place to grow up. We had a good high school—only one [laughs]; now I think there’s four or five in the area. I enjoyed it, even though I never wanted to go back there after I got out of high school. I went to Duke University and never really went back to Florida to live.

Abbate:

Were you interested in math as a child, or in science?

Willbanks:

Yes, I was. I always seemed to like those subjects better.

Abbate:

How did you end up at Duke?

Willbanks:

I don’t really know! [laughs.] I don’t remember. It just seemed like I had always wanted to go there. I didn’t want to go to school at a women’s college in Florida, and that was a place I had heard of, and it wasn’t that much further north of North Florida. I managed to get in and get accepted, and then after my first year I got a scholarship for the next couple of years—tuition and stuff—so it worked out fine. The price of going there now is about six times more than it was when I went! [laughs.] Just like any other college, I guess. It’s rather amazing how much it costs to send a kid to school—a good one, anyway—nowadays.

Abbate:

Did you major in math there?

First Job – Pratt and Whitney Aircraft

Willbanks:

I majored in math and physics. Then I got this job that I mentioned before the interview, with Pratt and Whitney Aircraft up in Hartford, and worked up there for two years.

Abbate:

Was that doing hand calculations there?

Willbanks:

Yes. The project was for a nuclear-powered aircraft.

Abbate:

Wow!

Willbanks:

Had you heard about that project?

Abbate:

I have, but it sounds insane.

Willbanks:

Well, there was a lot of work going on. Both Pratt and Whitney and General Electric were contracted to do feasibility studies. The main problem is shielding of the nuclear plant and all that, and so it never did come about.

I had a friend who married a guy from Texas, and they came out on a trip to the West, and I got these post cards of all the parts of the western United States! [laughs] And I was getting kind of tired of doing hand computing there at Pratt and Whitney, and one of the guys that I worked with there said, “Well, why don’t you check into Los Alamos and see if they might have some jobs out there?”

Abbate:

So this was sort of a well-known place for people doing calculations to work?

Willbanks:

Well, there were five or six different National Laboratories at that time: Argonne and Brookhaven, and there’s one in Oak Ridge, and the ones in California, and this one; and this seemed like an interesting place to start. So I came out for my interview, and I was offered the job, and three or four months later I was here! [laughs.]

Abbate:

And that was in ‘54?

Willbanks:

Fifty-four, yes. I worked two years at Pratt and Whitney and then came out here.

Abbate:

For Pratt and Whitney, how did that work? Did they give you a problem that you had to break down, or did they just sort of say, “Calculate this”?

Willbanks:

We had sheets of paper about three feet long, with calculation equation stuff at the top of each column, and we would maybe have variables down each end, and we would just calculate across the page for these engineers. They were checking heat flows and fluid dynamics and stuff, within the reactors in what might have been a nuclear power plant.

Abbate:

But you were just . . . .

Willbanks:

I was just a peon! [laughs.]

Abbate:

It sounds kind of tedious.

Willbanks:

Yes, it was.

Abbate:

What was your job title doing that?

Willbanks:

I have no idea! [laughs.]

Moving to Los Alamos

Abbate:

When you interviewed to come to Los Alamos, what did they want you to do?

Willbanks:

Well, actually I got started in a small group that had some calculations to run on the MANIAC computer, which you’ve probably heard about from Marge [Devaney]. It was parts of weapons computations: they were in just little isolated parts here and there, and so I think it was a combination of stuff on punch cards and in the MANIAC. The guy that was my boss at the time was here three or four years afterwards, but he quit and went to IBM.

Abbate:

Who was that?

Willbanks:

Willard Bouricius.

Abbate:

Oh, I don’t know that name.

Willbanks:

Well, he worked at IBM for a long time, and he got to be a pretty important person there. He was a—well, I wouldn’t say a “child genius”; I guess he was a man genius. He was a smart guy.

So then he left, and computers were coming in. I guess the first one was an IBM 701. Then I worked with another guy that’s also pretty famous in the nuclear science world. His name is Ted Taylor, and he worked for General Dynamics out in San Diego after he left here, and then he had some jobs with the Atomic Energy Commission, and I guess he wrote papers and did reports and stuff after he left here. But he was a weapons designer at the time here, and so I just did whatever he told me to do: some calculations on computers, some hand calculations; whatever needed to be done. And it just sort of evolved: as one person I was working for seemed to leave, then I fell in with some more people.

Abbate:

Was the first time you saw a computer when you came here, then?

Willbanks:

Yes. Other than the hand computers, yes.

Abbate:

And that was the MANIAC? Or was that an IBM by then?

Willbanks:

Well, the MANIAC was the first one that was here. I didn’t work on the MANIAC very long. I got into the 701, and I got really into working closely with bomb designers. They would do the equations, and then I would translate that into the machine language code that needed to be put in. It just sort of evolved from punch cards up to typing in stuff on the computers at the end, so it’s hard to remember exactly all the old steps you had to go through.

Abbate:

What did you think of the computers? Was it something you had wanted to do?

Willbanks:

Well, it just seemed like it was a lot more fun than punching hand cards. [Sometime in the 1950s], I was working with one guy doing some calculations for this guy in J Division—Ph.D. guy—and he fixed it up so I was able to go out to the test site when they were still shooting weapons.

That was the Nuclear Testing Ground in Yucca Mountain, but it was the Nevada Test Site in those days. I actually got to go with him into the shed where the bomb was, and I got to help him make some of the connections and stuff. He had an experiment—each engineer had their own little things going on, and I was able to get in and help him, and then I got to see it when it went off.

Abbate:

What was that like?

Willbanks:

Pretty exciting! [laughs.] Of course, you had to have your back turned in the very first part, because of the flash, and then you had glasses. It was all over pretty fast, but it was definitely exciting!

Abbate:

Was it scary?

Willbanks:

No, I guess I was too much interested in what was happening to be scared. I mean, it was a long way off, and of course it was under a lot of supervision. It wasn’t any of the ones where they had people out in the field in various places and stuff like that; it wasn’t one of those kind; but it was exciting.

Then I worked in the Weapons Divisions till the early seventies, and then the boss I had left. He and another guy had been working together, and the other guy got to be Head of the Computer Division in Los Alamos: Roger Lazarus. I don’t know whether you’ve . . .

Abbate:

I’ve heard the name, yes.

Willbanks:

Marge worked under him for a while.

So then I went over to the Computer Division, and stayed there until I retired, through a bunch of different projects; but I was on the last one for twelve years, I guess. The last project was the mass storage project. Well, actually I was on two of those, but the first one was pretty primitive! [laughs.] Have you heard anything about the IBM Photo Store?

Abbate:

I don’t think I have heard about the Photo Store. What was that?

Willbanks:

Well, they had little plastic cells, about three or four inches by two inches, maybe, and there were something like 35 millimeter frames on chips inside. Each one held, I think, 32 chips, with probably 32 frames on each one. You’d send data in, through various and sundry different mechanisms, from computer programs running on these IBM machines, and it would actually get chemically developed on these little chips, and all this data was stored away on those.

Abbate:

So it was stored as spots on film?

Willbanks:

Bits, yes.

Abbate:

I hadn’t heard of that. That was in the sixties?

Willbanks:

Let’s see: yes, it was early sixties, mid-sixties; along in there.

Abbate:

How was that, as a storage device?

Willbanks:

Well, it was all we had [laughs], besides disks. One of the storage systems that I was working on had disks that were about two-and-a-half feet in diameter. They were stored as bits on those, too; but those tended to crash fairly often, and then rescuing the data wasn’t much fun.

The evolution of storage has sort of evolved along with computers, because the more people were able to compute, the more data they were able to generate, the better and more storage capacity they were needing. At the time I quit, they were coming out with the 3490’s—the little square cartridges that Storage Tech has developed. Have you run into that?

Abbate:

They’re in Denver, right?

Willbanks:

Yes. Boulder has a huge complex out there. The storage system I worked on was on magnetic tape; that was before the cartridges came along. So they sort of evolved together.

The last twelve or thirteen years that I was working, I was more involved in the software development for storing the data than I was actually with computing itself. I was on the storage end, rather than developing the codes and stuff to run on the computers, as weapons codes and research stuff that they do here in Los Alamos.

Abbate:

So that would be a switch from applications to systems, I guess?

Willbanks:

Yes. When I left the Weapons Division and went over into the Computing Division, about the time I got in there was when they switched over from IBM machines to Control Data machines, and then the Crays came along; and during almost all of that time I got into the storage end of it. That was the project that Marge came into that I was already on. She may have told you. It was in the mid-seventies, I guess; something like that.

Initial Job Description at Los Alamos

Abbate:

In the earlier phase, when you were working on weapons applications, how did that work? Was it you and a physicist working together, or was there a larger group?

Willbanks:

Yes, we had a group of five or six people. The engineer would give us a bomb design—different components; what it was; what the composition was—and then we would set it up. The codes were already written by then; we’d already done that. We would input the data, and then out at the end would come dynamic flow of the weapons after the simulated explosions.

Abbate:

So you had programs worked up to model an explosion, and so you would just vary the parameters depending on the bomb design?

Willbanks:

Yes, the variables. The actual compositions: what materials and the dimensions of the design.

Abbate:

Did you need a lot of math to do that? Or was it more just coding?

Willbanks:

Well, it turned out to be more coding equations. We had different sections: the main program went through and got cross-sections for the neutron scattering and all the interactions, and so each material had its own cross-sections. I had quite a bit, at one point in time, to do with collecting all that data and getting it input into the proper tables and stuff, so that the information was correct. Then you had an input section, and then you had a different kinds of radiation section—was the stuff mixed up inside?—and all that stuff. A lot of different parameters went into the calculations.

Abbate:

Did you have different people in charge of different parts of the program?

Willbanks:

Well, no. Different people would get different weapons to run tests with. For a lot of them we already had the data from the weapons, now that they had exploded; though they were trying to check out various ways of theoretically making it agree with the experimental data. That’s the whole point of doing the simulations. You’d get a code that seemed to reproduce the previous results pretty well, and so then they’d take that code and see what would happen if you did this, you did that, and changed this, that and the other, to get different yields for different designs. The designs were, I guess you’d say, allocated to Los Alamos; some were allocated to Livermore, and each laboratory was responsible for certain weapons to produce. It became a lot more difficult after the testing quit.

Abbate:

When they did a test, were you eagerly waiting the results to see if your computations had been accurate?

Willbanks:

Oh, yes! [laughs.] Yes: our whole group had made some promise that this was what it was supposed to do. Sometimes it did, sometimes it didn’t. But it was the Ph.D. designer guys that were responsible. I didn’t have any real input into that, other than that they would say, “Go try this; go try that,” and so I would; and I’d come back and tell them what the answer was, and then they’d say, “Well, try some other thing.” It was a lot of computations to go on for one final design!

Abbate:

Did you celebrate if it came out right?

Willbanks:

Not here. They may have out there! I didn’t get to go out into the Pacific when they were having them out there. But I did get to go to Nevada.

Abbate:

What was the trickiest part of doing that kind of programming?

Willbanks:

I guess I don’t know, because when I first started out, it was all in machine language, and later, as they got more and more sophisticated software, it was a lot easier to do the programming. I guess I don’t know.

Abbate:

What kind of software was that, at that point?

Willbanks:

Well, most of the weapons design that I was working extensively was on an IBM machine called Stretch. It was one of a kind, almost; I guess maybe there were several built, but it was a unique machine, and so that essentially was a dead end, learning that.

Abbate:

Did it have its own language, or was that FORTRAN or something?

Willbanks:

No, it wasn’t. It was all in the machine language.

I guess it’s hard to say what the most challenging part was. That was so long ago now, it’s hard to remember exactly what all the hard parts were.

Abbate:

Was there something especially fun?

Willbanks:

Well, I guess it was just a challenge. It was sort of like a puzzle; a jigsaw puzzle or something. It was a challenge to make it do what you wanted it to—different ways of programming different equations, and stuff like that. I guess it was fun in the end, where you could get it to streamline the setup parts, and then eventually you got to the point where you could get graphics and stuff out, and finally got graphics programs and stuff that you could feed stuff into and get a picture, instead of having to look at large listings.

Abbate:

How did you learn to program?

Willbanks:

Just by being instructed by these people that I was working with. You just got started. You were doing essentially the same sort of calculations you had been doing on the hand computer, except now you were doing it with punch cards: you encode operations on punch cards, and then you get an output from this, and then you do something more to it. It was all in steps.

Abbate:

So it was really similar to hand computing?

Willbanks:

Yes. It’s the same basic equations and algebraic stuff to do things to compute pressures and temperatures and stuff like that, and then the end product of the whole thing was a yield: how many kilotons did this thing produce? That was the end result.

Abbate:

Did your physics background help?

Willbanks:

Well it certainly helped, because I knew what the terms were, what basically was going on inside one of the weapons; but once you get into the computing end of it, the physics part sort of goes away. You let the Ph.D. guys take care of that. They tell you what equations they want in there, and then you go and fix up a programming structure that was sort of like subroutines: you’d call one with certain inputs, and then something comes out, and within the subroutine is how they assume different things were going to interact inside; different theoretical equations.

Abbate:

It sounds like you were doing fairly structured programming, if you had subroutines.

Willbanks:

Yes. Yes, all that part was more or less laid out. Then when I got into the storage end of it, it was mostly dealing with input-output devices and getting blocks of data onto these storage devices; so that was, I guess, essentially structured as well, in that you had to learn all the particular characteristics of these storage devices and block the data accordingly, because certain hardware accepted different-sized blocks of data. So you had to figure all that out.

Abbate:

How was the storage system set up? Were there many different types of computers connected to the photographic device (or whatever the later ones were)?

Willbanks:

Well, there was an interface. The people did their code on the computers. They were all in a network. The computers didn’t talk to each other but only to the central filing system. They’d run their codes on one of the computers, and then they’d send the data over the network to our system, which would then block the data and write it out onto whatever storage device seemed appropriate. If it was a small file, it’d go on a high-speed disk; if it was a huge big file, it would go onto the slower devices, whatever they happened to be at the time: magnetic tape, or the newer tapes, or something like that. Photo Store didn’t live too long—I would say its lifetime was maybe seven or eight years—but at the end of that, when we had the new system, we had to write software programs to transform all that stuff over and save it into the new system. So there was a lot of stuff we had to write to manage to do that, because the users didn’t have to bother with that; we did it for them. We just told them where the new data was and what it was called, and they told us—I’ve forgotten now how we really worked all that out, but we didn’t lose it. They wanted it saved; we transcribed it onto the newer devices. For some of that data, it’s happened about three different times now, it’s been shuffled around. So some of that original weapons data is still out there—though not many people know what it is!

Working on Storage Systems

Abbate:

Now, you worked on storage systems for twelve years or something?

Willbanks:

Well, since 1979. Our current one that’s still running over there, the CFS Project, went into production in 1979, and it’s still running over there—part of it.

Abbate:

It sounds like a big project to have kept you occupied for so long. What were some of the challenging things about that? What kind of problems did you have to solve to do that?

Willbanks:

Well, you had the security aspects of it for the classified data, but the more challenging part was, every three or four years they’d come up with a new storage device—a faster disk; a better tape—and so then you had to go write the interface package to put that data out on the new storage device. So it was constantly evolving. And then the users would want different things available to them, to either protect their data or to share it with somebody else, and to make it easier and easier for them to tell us what kind of data it was, how valuable it was, and what they thought about it, so that we could decide better where to put it. It just kept evolving; and as each new computer came in, somebody on their system side had to put together an interface that interfaced with the user and then to us, so it was file transport codes and stuff like that that we had to work on. It was a lot of different things.

I wasn’t the only one. The file system occupied six or seven people, more or less, for ten or eleven years.

Abbate:

Did you have to write user interfaces to this as well?

Willbanks:

I didn’t work on the user interfaces. I worked on our interface to the piece of software on the computer that interfaced to the programmer.

Abbate:

Okay.

Willbanks:

The application program would say “save this file,” and there were parameters that they would give; and then they had a sort of a black box that took that information and put in the various parameters, like how big the file was and what kind of format it was; and then that came across to our interface, and then we decided what to do with it; and then we had the file transport at the lower level, just sending the box of data back and forth.

“Abbate:”

That was kind of a client-server system?

Willbanks:

Well, yes.

Abbate:

Did you have to handle simultaneous requests from different users wanting to save things?

Abbate:

Oh, yes! We have multitasking on IBM. I guess we fixed it up, eventually—that was one of the challenges, to get it to perform well—up to like forty tasks at a time, so we could be doing ten or twelve file transfers simultaneously. But unfortunately, all of the data still had to go through the CPU, and that limited the speed. And then there were the hardware pieces: the hyperchannel and various and sundry interfaces on each end. I had nothing to do with that, that I could avoid! [laughs.] The tapes and disks were bad enough for me to work with!

Abbate:

Was there nothing commercially available? Is that why you had to write your own?

Willbanks:

We were commercially available. [laughs.]

Abbate:

So you were exporting this to other places?

Willbanks:

Yes. That’s how I got into the consulting work with the Weather Center in England. Then I took a year off, and there was a commercial company called DISCOS that was part of General Atomics, and they were marketing the program under the CFS trade name. At one time we had up to seventeen against the license, and they were running it at San Diego (UCSD), they were running it up in Berkeley, and we had some at Idaho Springs, and we had it sort of running once actually at an IBM place. Yes, they sold it.

Abbate:

I’m surprised IBM didn’t adopt it, because—this was on an IBM system?

Willbanks:

Yes.

Abbate:

But IBM didn’t adopt it?

Willbanks:

No, they didn’t adopt ours. They have been marketing their own since I quit—trying to, anyway. But the market for this was limited; this was about the limit of how many installations had this kind of a need for that much data. People that were doing bomb calculations, or people that were doing weather, or something that generated lots of data; it wasn’t an on-line enough system for banks or commercial things to really be interested in. So the company, DISCOS, thought they could make money out of it, but . . . Savannah River had some for a while, and essentially all the AEC installations, but it wasn’t something that really filled a need, because about that time was when all the interactive stuff started coming in with the PCs, and all that sort of thing. For our laboratory, it still worked, but it never flew as a commercial product. We talked to the CIA about it. Where was the other place we went? Oh, down in Houston: we went down to an IBM installation down there, and they had an idea about doing this and that and the other to market it, but that didn’t quite work out either.

Abbate:

Would that be for the space program?

Willbanks:

No, not really. It just happened to be the place where this branch of IBM at the time had its offices.

Abbate:

Did you end up giving talks at computer conferences about this?

Willbanks:

No, I didn’t. The project manager did most of that.

Abbate:

Who was that?

Willbanks:

Bill Collins.

Abbate:

I don’t know him.

Working for the Weather Center in England

How did you end up going to the Weather Center in England?

Willbanks:

Well, they were looking for a storage system, and they couldn’t find one to buy, and so some of their people came over here and asked us if we could give them the software; and since it was essentially in the public domain, we said, “Sure.” So Marge and Bill went over there first and just talked to them about it, after their guys had come over here, and then we decided to go ahead and install it, so then I went over for a month or something and got it installed. Then they had to make a lot of local modifications to make it do what they wanted it to do, and we didn’t hear anything from them for some little time; and then they decided they wanted to upgrade and get into the newer storage devices. So I went over there maybe a half a dozen times altogether, to help them with it.

Abbate:

What was that like?

Willbanks:

It was fun! [laughs.] I really enjoyed that. I’m still friends with four or five of the people I worked with over there.

Abbate:

That was in Reading?

Willbanks:

Yes. Yes, it’s just about forty miles outside London, on the west side toward Wales.

Abbate:

And that was some sort of pan-European cooperative venture?

Willbanks:

Yes.

Abbate:

What kind of computer did they have?

Willbanks:

Well, at the time when we first went over there, I guess they had Control Data machines; and then the Crays came along; and then they had a Fujitsu machine, fairly recently; and now they just signed a contract to get an IBM machine. But they’re not running that software anymore. They quit probably three or four years ago. They finally got a replacement that was a commercial product, from IBM.

Abbate:

So I guess no one in Europe had anything like that.

Willbanks:

No. I was over there one time when they were having a conference at the Weather Center, trying to get together with other large installations: CERN people were there, and they were there from the British Met Office; they had five or six different installations that they tried to get together to write up specs for a system. I think they finally did get it together, but then they never got anyone to bid on it who really sounded like they could do what they needed to have done.

Abbate:

And you already had it.

Willbanks:

Well, they already had our product by then. Ours was what was running at the time, but because it was home grown, I should say, it took personal care in order to upgrade it. It wasn’t a black box, and they couldn’t have a vendor like IBM to tell them they want to do this or that; they had to do it themselves, and they didn’t like having to have their own personnel do the work. We didn’t mind doing it here; Los Alamos has done a lot of their own stuff over the years, because they couldn’t buy what they needed from software vendors. It had built-in security and stuff, which we needed and Weather Center had no use for. They have an IBM system now called HPSS, which is the only commercial product, I guess, of any magnitude, storage-wise. That is going with varying success, depending on various installations; but I guess there’s half a dozen, maybe ten locations running it.

Abbate:

Was that a very different environment to work in?

Willbanks:

Over there at the Weather Center? No, not really. They had very qualified people over there. All the people that I worked with were very sharp, and you didn’t have to tell them anything more than once; most of the time they understood. They were able to do what they had to do to modify it for their personal needs. They didn’t want to bite off, so to speak, the job of doing the modifications on their own for different hardware; so when we went ahead, then after we had it working they’d say, “Well, we’ll try it now!” [laughs.] So that was how that evolved.

Abbate:

Were there a lot of women there working with the computers?

Willbanks:

Over there? No, not really. They had some IBM people in there, and they had a few women meteorological people. Maybe a quarter of the people were women. It’s hard to remember exactly who was there when I first went over there.

Experience of Working on Classified Information

Abbate:

Was all of your work classified? Other than the weather stuff, I guess?

Willbanks:

Over here you had to have a clearance to work, even though we didn’t really actually deal with the data itself; we didn’t know what it was. But anybody that was in the secure area had to have clearances, and we were in a secure area, because that’s where the computers were.

Abbate:

I guess the MANIAC wasn’t in a secure area—the original one hadn’t been; but then the later ones were?

Willbanks:

Yes, I guess that’s right. Security was pretty lax, I guess, in the early days, and then as it got to be more and more of a reservoir for the weapons data, it got to be more and more of an incentive to protect it in various ways.

Abbate:

What was it like to work in a secure area? Were you always conscious of guards and badges and things?

Willbanks:

Well, the only guards were at the entrance to the buildings or the tech area; there weren’t any guards inside. Once you got inside, you wouldn’t know it was a secure area as opposed to a non-secure area, except that you did have safes in your office. But our code itself wasn’t classified, because it was only a storage system; it was the data that was in it that was classified. So it really wasn’t much of a problem for us. We weren’t directly involved in it, other than making sure that the data that came had to have passwords, and stuff like that. That was built into the system: if it was secure data, it had to have passwords, and only the user that sent it had access to it, unless they specifically put in other users who could access it: because a lot of the groups would have group codes, and each individual scientist would run different problems, but using the same weapons code, or whatever it was. So that was really about all the security involvement that I had.

Abbate:

When you were working on the earlier weapons problems, would that have been more classified?

Willbanks:

Well, that was the days where it was all on computer paper, and listings went into safes at night.

Abbate:

You put your programs in a safe at night?

Willbanks:

Not the programs, but the data of the problems. You know, you’d run a problem, and you’d get a stack of listings about three or four inches deep. Even though the columns weren’t labeled—you just knew from memory, or you had something that told you what all the columns were—it was stamped “restricted.” Almost all that was “restricted” data and not “secret,” and so it just went into a combination safe at night, all the listings.

Abbate:

Were you not allowed to talk about your work with people outside?

Willbanks:

That’s right; we weren’t. I mean, everyone knew what you did, but . . .

Abbate:

Generally speaking.

Willbanks:

Yes.

Abbate:

Just not the actual equations or something.

Family and Personal Life

What did your parents think of your work?

Willbanks:

They thought it was a good job, I guess. They came out to visit a couple of times. They lived in Florida, and my dad couldn’t take the altitude here, so mostly I went to visit them.

Abbate:

Had they encouraged you to have a career? Were you expected to support yourself when you grew up?

Willbanks:

Yes, I think I was expected to support myself! [laughs.] They were sufficiently well-off enough to send me off to college, but they never were rich. My dad was just a worker; neither of them went to college; and I was born at the beginning of Depression, so things were a little tough on them, too—but at least my dad had a job. I was never deprived of anything I really needed; but then, on the other hand, I didn’t have any luxuries. [laughs.] But I didn’t need them either; I didn’t care.

Abbate:

But they weren’t expecting you to just get married and be supported by someone else?

Willbanks:

Well, I guess either way they wouldn’t have cared! [laughs.] Whatever I wanted to do, I suppose, was all right with them.

Abbate:

That’s good.

Did you meet your husband at the lab?

Willbanks:

Yes. Yes, we were in the same work area. Not in the same group: I was still in the Weapons Division, and he was in the Computer Division. We got married in ‘59, and he died in ‘94. He got a very sudden brain tumor that was inoperable, and from first symptoms till he died was like from the first of December to the middle of February. It was very quick. No one understands how, but that’s the way brain tumors go. That made quite a difference in a hurry.

But he worked in the Computer Division, and he actually got to go over to Enewetak and Bikini when they were having the testing over there in the Pacific in the late fifties.

Abbate:

Was the lab a pretty social place?

Willbanks:

Yes, it was a fun place. There was a lot of interaction with your co-workers, and there were Christmas parties, Division parties . . . It’s a lot different now!

Abbate:

Oh, really?

Willbanks:

Well, it was when I left! I don’t know what it is now.

Abbate:

Why is that?

Willbanks:

Well, there were a lot fewer people then, for one thing.

I always have, I guess, socialized a fair amount with the people that I have worked with and under, and there wasn’t a lot of caste system. You know, you could speak to the Director of the Lab or the Director of the Division and your bosses, if they had parties and you went to them. It was not very structured. Everybody was just friends! [laughs.]

Percentage of Women Working at the Labs

Abbate:

Were there a lot of women in the Computing Division?

Willbanks:

Well, there were certainly a lot of them that were doing key punching. Essentially they were all women doing that—maybe there was one guy—because they were really sort of like secretaries and typists.

Abbate:

But were there women on the programming level?

Willbanks:

Yes, there were a reasonable percentage. I guess I don’t know what percentage; I never really thought about it. Probably a third, I suppose, of the people that were doing the computing were women. There were a couple of the Ph.D. physicist women that were reasonably well-known in their fields. There was no—I never, anyway, in all the time I was over there, felt any discrimination. I got promoted; I got a good salary; I don’t think it made any difference to anybody that I was a woman as opposed to a man. I did my job, and I got good raises, and I had nice bosses, too!

Abbate:

So you thought you were pretty much equally compensated, as far as you could tell.

Willbanks:

Yes. At least up until the time I left. I really don’t know much about what’s going on over there now.

Work Environment and getting a Masters

Abbate:

What was it like, working with the scientists?

Willbanks:

Well, since I was a pseudo-scientist, and not just somebody that had come there out of high school and sort of grown up in the system, why, I felt that I fit right in. As I say, it was a fun place! I think everyone enjoyed working there. We all felt that, during the Cold War and all that, that the work we were doing was important. I think that may be part of the problem nowadays, that we no longer have that feeling. We were all working toward a common goal. But as I said, there were a lot fewer of us. That was just in our division; I don’t know what things were like much in the other divisions, but I think even the whole Lab felt we were on a mission—and I guess we accomplished it. Afterwards, maybe the whole attitude of scientists sort of changed a little. It’s hard to say, since I wasn’t one: I got a Master’s, but I didn’t get a Ph.D.—not that that really mattered.

Abbate:

When did you get the Master’s?

Willbanks:

At UNM in ‘57.

Abbate:

University of New Mexico. Did Los Alamos sponsor you to do that?

Willbanks:

Well, you were able to take courses here, up until you needed to do a thesis; I had to spend a semester down there to do that.

Abbate:

That was the main campus in Albuquerque?

Willbanks:

Yes.

Abbate:

But they had a sort of extension campus up here in Los Alamos?

Willbanks:

Well, no, it wasn’t an extension campus. The people in the lab, the lab employees, taught the courses. Most of them weren’t instructors from the University; a lot of them were taught by people from here. There were some who came up from Albuquerque.

I guess that our classes were either in the high school or in the Lab property. There was no UNM campus—there was no UNMLA—in ‘57. They paid your tuition, and they cooperated, and they’d give a leave of absence when I needed to go down there to do the work. But there were a lot of people who took courses from the University up here and got higher degrees—or, if they didn’t happen to have a college degree when they got here, they could get one. Like a key-puncher: if they wanted to, if they were ambitious enough and could hang on long enough to take all the courses, then they could get a degree, and get some advancement at the lab because they got a degree.

Abbate:

So there were women who sort of worked their way up from key-punch to something else?

Willbanks:

There were some, yes. But there were more men who did that. Some of the guys that were in the machine room doing operations: there was this one kid that I knew that got out of high school and went through the whole process, and now he’s doing science work over there in one of the divisions. Several of them have done that.

Abbate:

Was it all men in the machine room?

Willbanks:

No, no. But it was the men that seemed to be a little bit more ambitious, from starting out from a high school level.

Abbate:

Were these local people, the high school graduates?

Willbanks:

Well, some of them were the sons and daughters of people that lived here already, and certainly several that I know were local; but then there was another guy that came from back East somewhere, and he went through the whole process. I think he wound up actually getting a Master’s. It was a place where you could advance, if you were conscientious and ambitious enough to do it.

Abbate:

Did you have any children?

Willbanks:

No.

Abbate:

So you didn’t have to worry about child care.

Willbanks:

No, I didn’t have to worry about any of that— or the schools here.

Abbate:

I hadn’t even thought about the schools here.

Willbanks:

Well, I think they’re pretty good quality. I think they may not be as good now as they were. At least that’s what I hear. It’s hard to get good teachers anymore, when you can’t pay them.

Technical Details of the Machines and Processes

Abbate:

Did you work long hours when you were there?

Willbanks:

Sometimes I did, yes: because sometimes you’d have deadlines, and sometimes it was easier to do work in the middle of the night; it was easier to get on the computers than it was in the daytime. There was sort of an allotment; each group would sign up for computer time, and the weapons groups had priority, and then there were other fill-in jobs. If some of the weapons programs would crash and wouldn’t run, then they’d run something else; and sometimes you’d get called in the middle of the night because the program that you had set up had crashed. [laughs.] And sometimes you’d go in, depending on what the circumstance was—whether you were supposed to have an answer for your physicists in the morning or not! [laughs.]

Abbate:

So that these programs would be running for hours?

Willbanks:

Oh, yes. Some of them take five or six hours. And then maybe you would look at where it was headed, and it didn’t look like it was going to work out, so then you’d take dumps as you went along, and you’d back up and change something, and then continue again. These computers weren’t very fast compared to nowadays!

Abbate:

This was IBMs, still?

Willbanks:

Yes, this was the 701, the 709; and in parallel to the 709, I think, was the Stretch computer that they developed, and it was an end within itself. Then they started the parallel computers, and that was a whole different ballgame from the applications side, but I was out of that by then.

Abbate:

Those were Crays, the parallel ones?

Willbanks:

Yes.

Abbate:

It seems like if something’s going to run for hours, you must have certain strategies in terms of—you know, you don’t want it to run for hours and then crash and lose all your results. Did you have to think about checkpointing and checking things out? How would you approach that?

Willbanks:

Well, you’d take tape dumps, back in the old days, every fifteen or twenty minutes or half an hour, something like that.

Abbate:

And those were so that if it crashes, you could read that back into memory?

Willbanks:

Yes. Because it wasn’t only the programs that would crash; sometimes the machines would crash. They don’t do that anymore, but they did then! [laughs.] They’d have a hardware problem, and it was gone; if you didn’t have a backup, you would have to start over. It got to the point where you’d automate it, and if the machine crashed the operators could figure out what your last dump was, and they would somehow or other tell it to continue on from there. If they had trouble doing that, then you might have to go in and straighten it out. I certainly spent a number of nights in the machine room! [laughs.]

Abbate:

And then also, if I understand, you would be maybe printing out intermediate values to see if it was doing what it was supposed to?

Willbanks:

Oh, yes. Yes, it was a continuous thing. Every ten or fifteen minutes there would be an output cycle, and that would tell you what things were going on. If you were there, then you could look at it; if you weren’t there, you’d look at it later, the next day when you came in. But most everybody’s philosophies were about the same: you didn’t want to waste more than a half an hour of computing time backing up—because it didn’t take long to take the dumps. You know, it was just a snapshot of what was going on.

Abbate:

And you could tell: “This doesn’t seem to be right, so let’s . . .”

Willbanks:

“. . .back up and change.”

Abbate:

Back up and change: so instead of reading the whole program back in, you were changing it on the fly?

Willbanks:

Yes, because they had . . . Well, it’s hard to explain, I guess, to somebody who’s not familiar with the weapons, but with any hydrodynamics problem, the way the fluid dynamics works, you can change the grids. When you first start them out, you’d have a coarser grid, and then as you get further into the things starting to expand and explode and all that, then you need to get more points in your grid. So often what we’d do, we would run it until it the geometry that we imposed on it, and the points that we were using to do the calculations of the pressures and the temperatures, were no longer adequate. You’d get to a point where you’d have to split it into a finer calculation, so that you had more points in there.

Abbate:

A higher resolution picture?

Willbanks:

Yes, exactly: that’s what it was. So then you’d go along for some period of time, and then that one wasn’t good enough, and so you’d have to keep breaking it up. But the more points you had, the slower the calculation, because instead of doing a hundred points, it would get to where you’d be doing calculations for a thousand points. You just had to use some judgment, and of course if you did it long enough, you’d have a pretty good feel for what you need to do when. You’d get familiar with that kind of stuff.

So that’s where you’d have to sometimes back up, because you’d essentially hit a dead end with this geometry; you had to go back and “split it,” is what we used to call it. From one point you’d go to two points, and then later you split it again—sort of like stock splits! [laughs.]

Abbate:

Would you develop sort of an intuitive feel for that?

Willbanks:

Yes.

Abbate:

It sounds like there was really quite an art to that. It’s not just mechanical; you might need a big-picture sense of the problem?

Willbanks:

Well, it was sort of a guessing game, and a challenge—to see if you could get the best result in the smallest possible time. [laughs.] Because that’s what the powers that be were really waiting for: “How did this come out?” So then you’d tell them, and then they’d go off and have a meeting and tell the Division Leader, and this and that and the other.

It was fun then, too, but I enjoyed the storage work, too. That was a challenge of a different kind. It was interesting.

Abbate:

It sounds like you’d have a different interaction with the users with the storage project, rather than sort of one-on-one with the physicists.

Willbanks:

Yes, those are completely different.

Abbate:

So the whole lab was your user base, in a sense?

Willbanks:

Well, yes. But most of the interaction with the user was with the people that wrote the black-box interface on the application side, to talk to the storage system. Users didn’t really interact with us directly; we would have meetings from time to time, and they would tell us, “Well, we would like to be able to do this,” or “We’d like to be able to do that.” But it wasn’t an every-day¬ type of situation. As long as we took their data and kept it for them, that was all they wanted to know! [laughs.]

Abbate:

Did you train anyone?

Willbanks:

Not really. We all sort of seemed to come in to it with about the same amount of experience, and you sort of worked in parallel, along different paths of getting different pieces of the code together at the end. So I wouldn’t say I trained anybody—except the operators in the machine room; but that was different.

Abbate:

Yes.

Was it difficult to coordinate that project, with ten people working on the storage?

Willbanks:

Well no, not really. It all seemed to work out pretty well.

Abbate:

That’s amazing! [laughs.]

Willbanks:

Yes. Well, we had pretty good people, and that was back in the days of structured programming and flow charts and all that, so everyone sort of knew where they were and where they were going, and we had code reviews and stuff like that. It all seemed to come together. In fact, we even got in production about a month ahead of schedule on that, so everybody was happy.

Abbate:

That is amazing, in my experience.

Willbanks:

Yes! [laughs.]

Abbate:

Why don’t they use flow charts anymore?

Willbanks:

I have no idea! [laughs.] I think it’s because they have such a high-level language—you know, FORTRAN and various and sundry of the higher-level languages—you don’t really need to go down into the nuts-and-bolts level like you did there.

Abbate:

So you think the structure is more in the code itself now? You don’t need to have it separately on a chart?

Willbanks:

Yes, I think that’s right. You do a lot more with—well, I don’t even know what they’re using now: UNIX?

Abbate:

Well, do you think it’s also true that because it’s easier to debug, people aren’t so careful when they’re first planning it out?

Willbanks:

Probably. That may have something to do with it.

Abbate:

It sounds like you had a lot more at stake in getting it right the first time. I mean, you didn’t have that much computer time to run it.

Willbanks:

Yes, we had to share it. The whole lab, everybody had their desires to get machine time! But it all seemed that it worked out.

General Thoughts about Working with Computers

Abbate:

Did you have any mentors at the lab, or role models? People you admired, or who helped you out?

Willbanks:

Well, I don’t know that you’d really call them mentors. The weapons guys—the two that I worked with the most closely—they were young Ph.D. genius types, and I guess in a sense they had to have been mentors, because I didn’t have all the physical knowledge. You know, they could generate these equations and all that, and I had no capability of doing that; but then, they weren’t training me to do that, either; I was down at a lower level. But they were nice guys. I guess those were the only ones, really. For the storage project, the guy that was the manager just got to be the manager by default. I mean, we were more or less equivalent, all of us in the project; it was just that one guy had to be the manager. So he wasn’t a mentor in any sense; he was a coordinator.

Abbate:

What have you found to be the most satisfying aspects of working with computers?

Willbanks:

That’s hard to say. I guess it was being able to make them do what you wanted them to!

Abbate:

Did you have a sense of power?

Willbanks:

Sometimes it was frustrating: to know what you wanted it to do and not quite get it to work out right. But there are a lot of different ways of getting from a start point to an end point, and there were a lot of different ways of figuring out how to get there, and I guess that was the challenge. And sometimes you could make it come out neatly, and sometimes there was only an awkward way to get it done, seemed like! [laughs.]

Abbate:

Were you trying to find the elegant solution, if you could?

Willbanks:

Well, of course!

Abbate:

Rather than just the first one that worked.

Willbanks:

Well, they were more streamlined; fewer operations. That was always a challenge, back in the days when you were counting operations, trying to figure out how to minimize them. You went this way or that way, and you tried to avoid divisions . . .

Abbate:

Let me just recap: You were talking about avoiding divisions, and I had asked if that was part of the art of it: that you had to think of tricks to avoid slow calculations like divisions.

Willbanks:

Yes. Nowadays no one would worry about that sort of thing, I don’t think.

Abbate:

Those tricks: did you come up with those yourself, or were they sort of passed around among the programmers?

Willbanks:

Oh, I think they were more or less standard procedures. I guess those were things that sort of got passed down from one to another: somebody figured out something and said, “Oh, yes! We’ll share that!” [laughs.]

Of course, now it’s a different story with the applications, because they are doing parallel processing and multi-processors and all that kind of stuff, and I really don’t know how that’s handled anymore, because I never had any experience with that. And the PC is a completely different ballgame.

Abbate:

You’re saying it’s a black box, and you don’t know or care about what’s inside?

Willbanks:

Yes. I think there’s more black boxes than there used to be.

Changes for Women in Computing

Abbate:

Do you think computing is more open to women than in your day, or about the same?

Willbanks:

Well, I think it was open to women then, just a lot of them weren’t interested in computers or math. I think nowadays women are probably a lot more aware of what they can do, and there used to be, before the last year or two, a lot of good jobs in computing for both sexes. The whole new Internet opened up thousands of jobs to anybody; but now, I guess, a lot of them are being cut.

Abbate:

When you were in college, majoring in math and physics, was that pretty unusual? Were there a lot of women in your math and physics classes?

Willbanks:

No. No, I was the only one in my physics class, and there were four or five, maybe, out of twenty-five or thirty in a math class. I was the only woman physics major in my class at Duke. We only had two hundred girls, or something like that; it’s not a big school, as far as undergraduates are concerned. There were a few women in the science classes in high school. I think a lot of women, at that time I came along, just weren’t interested in math and computing—well, computing didn’t exist; but in math. I think once they started having degrees in computer science at the universities, people got a lot more involved in learning. I don’t know.

Abbate:

Well, did you get encouragement from your professors to be doing that? Or did you just go your own way, regardless of what anyone else thought of it?

Willbanks:

Well, I guess I got encouragement from the staff at the university. I mean, they certainly were helpful, and they did everything they could to make sure you got through it and got a job after you got out. I guess I had never really thought about it.

Abbate:

So you don’t think they were discouraging to women, but just that women didn’t seem interested in doing that?

Willbanks:

Yes; yes. Because even out of my high school class, which had more kids in it than my college class, there weren’t a whole lot of men or women that wound up in computers or anything like that. A couple of them became doctors, and a lot of them became lawyers and businessmen; but there just wasn’t the same availability or interest in going that way, right at that particular time.

As the computers developed, then the interest in them did as well. I was right in on the ground floor. I think it took ten or fifteen years before people really became aware that that was a good field to enter, and then the whole thing just mushroomed.

Abbate:

Do you have any advice for young women thinking about going into computers today?

Willbanks:

I think it can be a fun thing. It can be a challenge. If you get in the right places, it can certainly be good financially. (At least, it was up until just now.) Certainly ten years ago that was one of the highest-paying jobs, coming out of universities: good computer people. There always have been jobs.

I guess that’s about it! [laughs.]

Abbate:

Well, it sounds like you got a lot of satisfaction out of it.

Willbanks:

Yes! It was a worthwhile career, I thought, as far as I was concerned. I met a lot of good people, and I still have a lot of friends that I’ve kept over various times. For me, it was fine!

Abbate:

Well, thank you very much!

Willbanks:

Well, you’re welcome. For whatever it’s worth! [laughs.]

Abbate:

It was fascinating.