View source for Oral-History:Carol Shanesy ← Oral-History:Carol Shanesy You do not have permission to edit this page, for the following reason: You are not allowed to execute the action you have requested. You can view and copy the source of this page. == About Carol Shanesy == Carol Shanesy was born in Evanston, Illinois, in March of 1937. She grew up in Evanston went to college in Ohio at Denison University. She completed her Master’s in Mathematics at Northwestern University. After graduation, she spent most of her career at IBM at the IBM Research Computer Center in Westchester County, New York, part of IBM's Research division. There she worked on a variety of programming projects, under the direction of Ralph Gomory, for a variety of research projects for IBM's research mathematicians, physicists, and chemists. She left IBM for the Rand Corporation in 1969 where she worked programming projects for the New York City Fire department before returning to IBM in 1972 as a Systems Engineer, initially in the Public Sector group, working on programming projects on behalf of IBM for clients, chiefly New York City. In this interview, Shanesy reflects on her career at IBM and Rand. She discusses the many programming projects she participated in while at IBM and Rand. She also reflects on her career as a woman in computing. == About the Interview == CAROL SHANESY: An Interview Conducted by Janet Abbate for the IEEE History Center, 1 August 2001 Interview # 614 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, IEEE History Center at Stevens Institute of Technology, Castle Point on Hudson, Hoboken, NJ 07030 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: Carol Shanesy, an oral history conducted in 2001 by Janet Abbate, IEEE History Center, Hoboken, NJ, USA. == Interview == INTERVIEW: Carol Shanesy<br>INTERVIEWER: Janet Abbate <br>DATE: 1 August 2001 <br>PLACE: Carol Shanesy’s office at IBM === Background and Education === '''Abbate:''' This is an interview with Carol Shanesy, on August 1st, 2001. I always start by asking when you were born and where you grew up. '''Shanesy:''' I was born in Evanston, Illinois, in March of 1937. I grew up in Evanston, and I went to college in Ohio, and came back for a year at Northwestern University (which is in Evanston) and got a Master’s. '''Abbate:''' What did your parents do for a living? '''Shanesy:''' My father worked for the Department of Agriculture, and my mother worked, when I was little, on and off for different doctors. She was the Evanston substitute secretary. All of the doctors’ nurses and secretaries would arrange their vacations so that Mother could cover for them. Later she worked for Northwestern. Actually, she worked for a book company first, Kingsport Press, while I was in high school; and then she switched to Northwestern, and worked for a very unusual professor there, Dr. Abe Charnes. Have you ever heard of him? Operations research. He’s now at Texas, I think; or he’s probably retired. '''Abbate:''' So, did you hear about this as a child? Something about O.R.? '''Shanesy:''' Well, of course, by this time I was in college, so I wasn’t a child; but yes, I did hear about her adventures with the students. Dr. Charnes used to do a lot of consulting for companies because of his field, and so I would hear about people coming in and interviewing. '''Abbate:''' Did you have brothers and sisters? '''Shanesy:''' No, I didn’t. I was their only child. '''Abbate:''' And did your parents encourage you to have a career? Did you think that was taken for granted? '''Shanesy:''' I think from a very early time it was a given that I was going to college. I did well in school, and I understood—possibly later—that one of the reasons is that my father didn’t go to college, and felt it handicapped him, and regretted it very much. He was really a self-taught person. He was quite smart. So he was very keen that I should go, and I never resisted in any way; I wanted to also. But I certainly entered college without any clue about what I was going to do in life. I think that’s a little different from nowadays; I think that career choices have moved up a little bit. But in my day, it wasn’t unusual. First of all, there weren’t nearly as many things open. So I just went to college, and waited until I was a sophomore to choose a major—maybe a junior; I don’t remember. '''Abbate:''' When you say there weren’t as many things open, what do you mean? '''Shanesy:''' Well, girls were expected to teach mostly. They were to have families. My particular family didn’t put any pressure on me at all in that way, but I was aware of what was going on around me. Interestingly enough, about two years ago we had a fortieth reunion at my college, and I was editor of the paper, so I was “volunteered” to make up a newspaper, and go back and get old articles and see what was going on in those days. One of the interesting things I found was that in the Winter to Spring issues, there would be articles about recruiters coming to the campus, and what struck me about them was that there would be a whole list of companies—Ohio had a lot of chemical companies and engineering-type things—and it would say “men only,” “men only,” “men only,” “men only.” And then there would be one or two that were for women as well as men, but they would almost always be teaching or something somewhat clerical. It didn’t mean anything at the time, because it was what we expected, I think; it didn’t seem odd at the time, or I don’t remember it seeming odd, but there were very, very few listings that said they were interested in women. '''Abbate:''' So the companies would actually specify . . . '''Shanesy:''' Yes. They would specify men and women. Now, of course they wouldn’t do that today—but they certainly did then. '''Abbate:''' That must have become illegal, the separate job descriptions. I should pin down when that happened. '''Shanesy:''' Yes. But I don’t think we thought anything of it. '''Abbate:''' Now, you majored in math in college, right? '''Shanesy:''' Right. '''Abbate:''' Did you have a very early interest in math or science? '''Shanesy:''' I suppose I always found it easier than other subjects. I did find it easier; I was more comfortable with it. I remember when I was a junior in high school, they offered a speech class. You had to take English, four years, but they offered in addition an English speech class and an English journalism class; and since I was terrified to get up in front of anybody, [the speech class] was certainly out, and I hated to write as well, so I thought anybody would be out of their gourd to sign up for one of these things. I made the mistake of mentioning it to my father—pretty much in jest—and he, in a very uncharacteristic reaction, spun around and said, “You’re not going to make the same mistakes I made! You’ll take one or the other!” And so I took the lesser of the evils, which was the journalism, and learned a tremendous amount. I went to a good high school, but this fellow sort of taught you the mechanics—getting started, organizing, and so on—so I didn’t have troubles writing thereafter. That was a very lucky break for me, because also I met most of my close high school friends in this class. So it was one of those quirks of fate! '''Abbate:''' Was that something you needed later in life? '''Shanesy:''' You need it constantly! In fact, one of the things that makes me a little weird in my profession is that I don’t mind writing as much as my colleagues. I mean, programming and writing skills generally don’t go together—and I’m not a good writer naturally—but I do know the mechanics, and I do a lot of work on manuals and stuff like that. So it was a great boon. And it was good in college—I mean, you had to write papers all the time, and I had an easier time, I think, than some. '''Abbate:''' So that’s a skill that computing professionals need, but isn’t necessarily associated with it? '''Shanesy:''' I think you tend to be weak in it; and maybe the schools contribute to that, too. For all the English I had all those years, I was uncomfortable with it and not very good at it. In one year, this teacher taught a lot, and then I had another teacher in my senior year who was into poetry and the more creative aspects of writing, and she tried to undermine everything that he’d taught us! But it was a good combination, because she didn’t want us to be formulaic; she taught us the other side, as best you could teach that. I don’t think you can teach creativity, but you can get what creativity is there out on paper. '''Abbate:''' How did you choose Denison University? '''Shanesy:''' I needed a scholarship, and I wanted a small university. I grew up in a big town—it’s not big by New Jersey standards, but it’s big—and I went to quite a big high school, twenty-five hundred students or something like that, and I just wanted something small. So I applied to a couple of colleges, and I was sort of late—as I am in most things in life!—and I missed some of the deadlines; so I applied to Denison, and Carleton, and Ripon, I think—anyway, small Midwest colleges. Denison gave me a nice scholarship, and so I went there. I liked it. I loved it. It’s a great place. I entered in ‘54, and left in ‘58. '''Abbate:''' Was there any computing on campus at that point? '''Shanesy:''' No, there was no computing. '''Abbate:''' Few places would have! '''Shanesy:''' No. A friend of mine and I had sort of heard about programming. My father, as I say, worked for the Department of Agriculture, and while I was in college, they had just started to automate some of the work that they did. This was the Commodity Credit Corporation that keeps track of all the grain that’s stored all over the country. You know how the market supports work: they store the grain on farm, and then they store it in the town, and it has to be kept track of. They had gotten an IBM 650, and this sea of key-punch operators, and my father was just sort of interested in the whole process; he would have loved programming if he’d been at the right age. And so a friend of mine (who was also a math major) and I sort of had our ears open for programming. That was the word in use, and we were math students, so we got that mixed up with linear programming: I mean, two completely different things! [laughs.] We didn’t even know the difference at that time! There was nothing at Denison, and no thought of it at that time. Northwestern had a 650, I think, and I took one course [there] because I needed some credits, and we coded in machine language—and I mean machine language. Do you know the 650 at all? '''Abbate:''' I don’t personally know the 650, no. '''Shanesy:''' First of all, it was bi-quinary; it had this two-and-five number scheme? '''Abbate:''' Bi-quinary? I’ve never even heard of that! '''Shanesy:''' Well, that’s what it was. It had a drum—the memory was a drum—and it was a two-address system, so you had an op code and one operand, and then the second operand was where you were going to go next; so you bounced around the drum for efficiency. You didn’t go from instruction to instruction, or from location to location. [laughs.] So I can’t remember what program we wrote—I mean, it was probably to add two numbers; just the basics. === Attending Northwestern === '''Abbate:''' So this was after you got your math degree at Denison, when you went for a Master’s at Northwestern in math. '''Shanesy:''' In math, yes. My advisor in college was very keen that I should go to graduate school, but she wanted me to teach, and I knew I didn’t want to teach. But I thought, “I’ll go for the Master’s; that won’t hurt anything.” I got a little money from Northwestern, and I could live at home, so there was nothing terribly important about that. But during that year I realized that I didn’t want to go on to get a Ph.D. in math. I wasn’t very well prepared for Northwestern, and Northwestern’s advising was dreadful. I mean, I had had a course at Denison called “Analysis,” and when I got to Northwestern they said, “Oh! You’ve had Analysis? You’ll go into second-year Analysis.” Well, the content of the first course was completely different, and I had a terrible time doing it! But in the end it caused me to get a job at IBM! [laughs.] '''Abbate:''' Let me back up a bit. Your advisor at Denison wanted you to teach because you were a woman with a math degree? '''Shanesy:''' No. My advisor was a woman, [but] if that was part of her motive, I didn’t pick up on it at the time. We were close as an advisor and a student, and liked each other, and I think she just thought I had what it took and that I should do it. She enjoyed teaching very much, but I was reasonably sure even at the time that I wouldn’t enjoy it. I know I wouldn’t have, as a matter of fact; it’s not my thing. I have nothing but admiration for teachers, but it’s not what I like to do. '''Abbate:''' Did you have an alternate plan? '''Shanesy:''' No, not a very clear alternate plan. I had done some almost statistical work in the summer for the Department of Agriculture, where I worked because of my father. I mean, I had to go through the regular tests and everything, but at least I knew they hired students. The first few summers we just typed, and then they let me do some other things. So I thought maybe I could get into something like that, but I had really not much clue about where I was going to apply for a job. <br> === Recruitment by IBM === '''Abbate:''' How did you end up getting that first job at IBM? '''Shanesy:''' This is weird! As I said, I had trouble with this one course, and the professor for the class was always very helpful, and he said, “You can do it,” and “I’ll help you,” and so on; so I was in his office fairly frequently, especially at first, as I caught up from this year I had missed. It was a funny, long, narrow office, and you had to walk by another professor to get to mine; and so I knew that other professor, at least by name and sight, and he knew me. In the third quarter—Northwestern was on a quarter system—I took Complex Variables, because I needed some credits and it sounded interesting. But Complex Variables is almost exclusively the province of electrical engineers. That’s what imaginary numbers are useful for; I mean, they really occur in engineering theory. It was a huge class; it must have been a hundred and twenty students in one of those great big classrooms, and about the second day of this class, the professor said, “I want to see so-and-so and so-and-so after class.” And I thought, “Well, what did I do?” You know, it’s only the second time of class. <br> '''Abbate:''' You being one of the people he wanted to see? '''Shanesy:''' Yes, I was one of them. I recorded the fact that it was all of the graduate students that he was picking on. So we went up there after class, and he said, “Please do me a favor. I am so embarrassed,” he said, “I allowed the recruiter to come from IBM, and he’s been here since eight in the morning”—it was about ten o’clock by now—”and nobody’s been to see him, because I forgot. He just came, and I hadn’t announced it. Go up there and act like you want a job!” [both laugh.] So there were four of us, and we did! At some point in this interview I’m getting increasingly uncomfortable, because he wants me to come out to New York—this was for IBM Research in New York—and I’m [thinking], “This whole thing is a charade!” I didn’t ever tell him the truth, but what I told him was: I didn’t really understand it was in New York, and I would be wasting his money, and I really shouldn’t accept his invitation. And he insisted! I made it very clear that I probably wouldn’t take the job, and he said, “That’s all right. I think we’d like to have you. Come on out and take a look.” And I took the job! That’s how I got my job. Isn’t that awful? [laughs.] '''Abbate:''' No, that’s a great story! [laughs.] '''Shanesy:''' It’s bizarre. And it’s absolutely true. '''Abbate:''' So what convinced you to take it? '''Shanesy:''' It was a programming job. They took us around Research, and showed us some of the things that they were doing, which included little to do with hardware, but there was an experiment going on with how crabs’ eyes worked, and there was a lot of mathematical stuff. The unit they were hiring me into was a very small unit, probably ten or twelve [people], that did programming for mathematicians and some other types of scientists—mostly mathematicians—so it was a lot of applied math, and they wanted you to have a math background so that you could understand what you were doing. You just got assignments, and you worked with whoever it was and did their program for them. Of course this was all batch. It sounded interesting, and the pay—which, if I remember correctly, was eighty-three hundred a year—was more than double what my friends were getting. My last couple years in college we had a suite together with six girls, and I think without exception, all of the other girls taught. Of course, they didn’t work twelve months a year; they worked nine; but the salaries ranged from three to four thousand dollars. '''Abbate:''' For college teaching? '''Shanesy:''' No, high school and secondary. These were regular [college students]; these were not graduate degrees. It sounded awfully good; it sounded interesting; and I thought, “Well, I will try it! Maybe I’ll like New York.” I wasn’t very keen on moving away from home, because my parents and I always got along very well, and I didn’t know anybody in New York. '''Abbate:''' Now, where in New York was this? '''Shanesy:''' This was Westchester County, just North of Ossining, which is about 30 miles north of the city. It was at what is (or was recently) the Hudson Institute. It’s called Lamb Estate; it was this little cluster of buildings on a country road. '''Abbate:''' So you weren’t in a big city. '''Shanesy:''' Oh no, it was small . We were not in any town at all, and Ossining is a town of maybe 30,000. The next one up is Croton (where I now live, by accident!), and that’s seven or eight thousand. I mean, it’s really small. === Working for IBM<br> === '''Abbate:''' Now, this is the IBM Research Division? '''Shanesy:''' Yes. Most of IBM Research was in Poughkeepsie, but this was a different group, and they were down in Westchester. '''Abbate:''' And what was their role? Were you working for clients, or internally for IBM? Or doing basic research? '''Shanesy:''' At that time, it was pretty much basic research. As I said, there was this fairly large project having to do with how crabs’ eyes worked. '''Abbate:''' Why? '''Shanesy:''' What IBM’s interest in it was, I don’t really know, but they may have had some interest in recognition; it’s recognition systems that it’s most closely related to. '''Abbate:''' Some sort of visual processing? '''Shanesy:''' Yes! I don’t know whether it was crabs—I think they were crayfish, which are probably different—but they’re one of those creatures whose eye is made up of all these little cells that are independent, and the researchers were interested in how they process the stuff. I worked with mathematicians, primarily Dr. Gomory early on. He was a fairly well-known mathematician who became head of Research eventually: Ralph Gomory. He was a pioneer in integer programming, which is an integer solution to a linear programming problem. Do you know what linear programming problems are? '''Abbate:''' No, you’d better explain. '''Shanesy:''' Well, it’s when you’re mixing grain, or cutting up pieces of something. You have a bunch of requirements, and you have certain inputs to put in, and you try to optimize something, usually the cost of the output. He was a pioneer in problems of that sort where you couldn’t, say, use a “pound and a half” of this; it had to either be one pound or two pounds, in other words. Say you’re cutting rolls of paper, where you have ordered sizes of rolls of paper, and you’re going to cut them from a long roll. You can’t arrange it so that you get half of one of these, because they don’t want it half-size; they want one of the whole size. So the solution to the problem has to be all integers, and that makes it very irregular and much more difficult than a normal linear programming problem. Gomory worked out algorithms for solving those problems, and I programmed them. He was a neat fellow, and it was fun to work with him. '''Abbate:''' Was this research center possibly akin to Bell Labs or [[Xerox Palo Alto Research Center (PARC)|Xerox PARC]]? '''Shanesy:''' It was very akin to Bell Labs. Xerox, really, I don’t think was so much on the scene at that time. In fact, somewhat later, when we were developing an operating system—I can’t remember the exact details, but there was sort of a joke about competition between Bell Labs and us. Yes, they did very basic research. Theirs was of course geared toward switching technology and communications, but they did some quite basic stuff, and so did we. '''Abbate:''' How were the problems chosen? '''Shanesy:''' The mathematicians chose their problems to work on. Certainly integer programming has loads and loads of industrial applications. There were other people that worked on things like sparse matrices, which could be a linear programming problem where there are very few elements. A linear program problem is lined up as a big matrix, but if you have very few elements, then you can use other techniques to solve it, and there were people that specialized in that. The integer programming programs we actually published. In those days, programming was considered an extra little frill that you might give customers, so we didn’t charge for it, and published them at that time. Actually it was a long time before IBM billed for any program, and when it came to the operating system, that was free even longer. Those huge operating systems were free for a very, very long time. And the source was free: you could look at the source. Of course, when other people started— Hitachi and all the others, you know—they used our operating systems, so we stopped doing that at some point. '''Abbate:''' I hadn’t realized that the clients got the source to the operating system. '''Shanesy:''' Yes! You could look at how it was coded—if you were feeling really strong! [laughs] '''Abbate:''' If you really wanted to! [laughs.] '''Shanesy:''' But in the early days, it wasn’t as robust and as complicated as it is now, and people changed things. You know, they’d have a particular requirement for their installation—or they’d think they did—and they’d change it! '''Abbate:''' The source was in machine code, or assembler? '''Shanesy:''' Assembler language, yes. I never coded machine after Northwestern [laughs], although we used to patch in machine language. I think nobody could understand this now, but when you assembled a program: I told you that we worked in Westchester, but the computer was still in Poughkeepsie; so twice a day, a station wagon would make a run up with all the decks [of punched cards] to run on the machine. So you [only] got two runs a day, and if you had to reassemble your program, it was a separate run; you couldn’t assemble and execute back-to-back like you can now. So first of all, we desk-checked like mad! We wouldn’t want anybody to do that now, because it’s too labor-intensive. '''Abbate:''' Desk-checking means you sitting down and . . . ‘‘‘Shanesy:’’’ Looking and making sure you didn’t make a typo; that you didn’t punch the card wrong; that all the commas were in place. All the things that diagnostics tell you right away now, we tried to find out beforehand, so that we’d get a clean assembly the first time (or the second or third). And then when you’d find a bug, you’d patch it! You always had a little area of storage at the end of the program where you could add instructions, and you would branch out of the bad code down to the patch area and then do whatever it was, and then branch back! '''Abbate:''' Really! '''Shanesy:''' Sometimes we just patched with a little piece of red tape, right over the hole in the card; and then later we had something called an octal correction loader, where you could enter the op code in symbolic, but all the addresses had to be numeric. '''Abbate:''' Wow, so you actually . . . Now, when you did a patch that way, did you eventually fix the source? '''Shanesy:''' Eventually, when you got it all done and you had it [correct], you’d reassemble it and retest it, just to make sure you didn’t mess something up—which you usually did!—and then that deck was the program. It wasn’t down on disk some place; they loaded the deck! '''Abbate:''' Right. But temporarily, if you patched it you wouldn’t need to reassemble it next time you ran it. '''Shanesy:''' Right. I mean, people would have an assembly deck in a drawer [gestures as if holding a large stack of punch cards], and for a big program there would be a foot of binary and a foot-and-a-half of correction cards. '''Abbate:''' Really! '''Shanesy:''' Sometimes, yes. We didn’t recommend that, but people did it! [laughs.] I remember, after I’d been there a while, I sort of ran the help desk. The help desk in those days: I don’t even think there was a sign, but people knew that I’d been around a while and could fix things, and so they would just come into the room with their problem, and lay it on your desk. And this one fellow used to show up with this drawer of cards [laughs]—it was our payroll program, as a matter of fact—with this wad of correction cards in it! And after you put that many correction cards in, you’re afraid to reassemble it, because when you’re doing everything in absolute addresses, you make mistakes transcribing it back to symbolic. The help desk wasn’t twenty-four hour availability, believe me! '''Abbate:''' Were there a lot of women at IBM Research? '''Shanesy:''' No. There weren’t a lot of women at IBM Research. In fact, probably our group had more than most. I don’t really remember how many we were, because it kind of ebbed and flowed, but I think when I was first hired we were perhaps half women. But there weren’t women mathematicians. There really weren’t many women around, and what women there were were programmers. There were secretaries and so on, but there weren’t too many professional women at all. '''Abbate:''' So when you say “mathematician,” you mean someone with the title of Mathematician? '''Shanesy:''' Ph.D. mathematicians who were research mathematicians in the Math Department. '''Abbate:''' And they were the people who would set the problems? '''Shanesy:''' They were; and then we had a Physics Department, and they gave us a lot of work. The physics people worked on cryogenics in those days, and all kinds of things. Eventually it got to be some of the chip technologies; that was later, of course. What other departments did we have? Eventually we had what would now be information-processing type things, but mostly it was physics and math in those days—and chemistry. We had a lot of chemistry work, too. So, all of them would set the problems. '''Abbate:''' Now, you spent several years: you were there in the Research Computing Center from . . . '''Shanesy:''' ‘59 to ‘69: I was there ten years. '''Abbate:''' . . . and eventually you were supervising and managing. '''Shanesy:''' Yes. '''Abbate:''' What were some of the most memorable things you worked on there? '''Shanesy:''' When we started out, the way you ran a job was: you had a deck, and you put it in a drawer someplace, and the operators at their leisure would run it for you. So we really worked on turnaround, because that was a big productivity issue. You had to stop working on the problem until it had been run. You always had to have at least two things going, and usually three or four or five, because you had to wait until that run got back—and if it was a complicated run, you might have to wait overnight, because they would only do short runs during the day. '''Abbate:''' So you would prepare one problem, and then send it off to the operators, and then turn your attention to another one. '''Shanesy:''' Yes, absolutely. And since it was a productivity issue, we worked very hard at getting turnaround down to about two hours, which was considered very good in those days. During this time, more and more programming was going on, and there was more and more demand for programming. There was a shift: as a group, we tried to get the mathematicians and physicists and so on to do their own programming. We would help them a lot, but if they could do their own, they liked that, because then they didn’t have to come down to you and say, “Well, I want this little change made.” It was good all around. We didn’t force it; if they wouldn’t do it, we’d do it; but a lot of them found that they got much better service that way. So that let us turn our attention to getting the turnaround as good as we could—because they didn’t work on multiple problems at once. You know, they would go do something else, but it was even more of a problem to them. Then, during that time, we got to the point where we could—we were still doing batch, but we had ways of sort of loading things on and getting them through very quickly, so we’d get the turnaround down to even less than that. In order to do that, we changed the way we handled job input and output. The first step in running jobs was to make an input tape out of them; so there was a queue as you waited to get on the input tape, and then you would wait for the entire input tape to be run before anything else would happen, and then you waited for the output tapes. There actually were two to three [types of output]: you had printed output, punched output from programs, and plotted output in many cases. All those tapes had to be run through before you could get your output. So two hours was actually very, very good. That was before disks—before we had them available, anyway. When we got disk technology, we could put a job onto the disk, and queue it up for execution and then, as soon as it ran and produced some output, we could begin printing that output—but we didn’t have any operating systems that did that. There was a fledgling one that IBM had written, but we didn’t have the right equipment for it. We were a bit like shoemaker’s children: we got equipment last. If there was something like a disk that all the customers wanted, it was hard for us to get it, because of course you wanted the customers to get first choice—and certainly the marketing reps wanted the customers to get first choice! So we wound up writing an operating system for a coupled 7040/7090 system that we could do this with, and we had this big disk that we stored things on. And so we wrote that: two of us, primarily. A fellow named Peter Markstein, who was really a very, very smart, good guy (I think he’s at the University of Austin now): he and I, and some other people under our direction, wrote a little operating system to do this. And that was, I think, one of the most fun things I did in Research. '''Abbate:''' Do you know when that was, approximately? '''Shanesy:''' I can actually tell you: it was ‘64 and ‘65, because I had signed up for a trip to Europe that the IBM Club was sponsoring, and of course I was going with some of my friends—and this project was in trouble! We had a terrible bug, and Dr. Gomory, who was my manager’s manager, had told me I couldn’t go. He didn’t like to do that; that was really out of character for him. I guess it was Thursday night I found the bug. Nobody was around; Peter was on vacation or something. I was so excited I found it, you know! And we got it straightened out. '''Abbate:''' So you got to go to Europe? '''Shanesy:''' Yes. Friday night Dr. Gomory told me I could go to Europe—on Saturday! And I went. '''Abbate:''' Well, that’s motivation! [laughs.] '''Shanesy:''' Yes, it was! [laughs.] And I’m sure that I went in ‘65; I’m pretty sure. '''Abbate:''' Well, certainly operating systems are a big help to the programmers on the ground. '''Shanesy:''' Yes. Of course, by today’s standards this was pretty minimal, but it did the job. '''Abbate:''' What kind of background did you have in operating systems? '''Shanesy:''' In those days, you just programmed. We didn’t have any [background], much. We had gotten a little skeleton [of an operating system] from a professor at the University of Chicago, who told us that he had a working system and we could just adapt it; all we had to do was change a couple little things. Well, when we got this system, it turned out to be vaporware! They hadn’t quite written it yet! So we wound up writing 99 percent of it. And in fact, this bug was in the very small bit of code that they supplied: it was in the interrupt handler; the interrupt handler was getting interrupted before it was finished storing things; and that was actually their code. But we probably would have done it wrong, too; I don’t know. But at any rate, there wasn’t that nice distinction at all between systems and applications, back then. In fact, I actually started in [[FORTRAN]] first. No, that’s not quite true: I wrote an assembler language program first, but then they wanted me to learn FORTRAN, which was very new at that time. We did a lot of our programming in FORTRAN, because it was very well suited to the kind of programs that we were doing—particularly because we changed them all the time. Dr. Gomory would have an idea for an algorithm, and we’d try it out, and we’d run some tests, and he’d say, “No, let’s try this; let’s try that.” So we were constantly changing things. That happened in physics and chemistry also. '''Abbate:''' And FORTRAN is designed for scientific and mathematical applications. '''Shanesy:''' Absolutely, yes. I haven’t looked at FORTRAN for 30 years, but it was a very nice language; it was aesthetically nice. It may not have been syntactically perfect by today’s standards, but it was an elegant, natural language. '''Abbate:''' I think Fran Allen had made some comments about FORTRAN . . . '''Shanesy:''' Of course, Lois [Haibt] worked on the original FORTRAN, with the inventor, John Backus. '''Abbate:''' That’s right. I will ask her, when I see her tomorrow—ask her for the war stories. '''Shanesy:''' Oh, I’m sure she has wonderful ones. '''Abbate:''' Did IBM keep you trained on all the latest stuff? Did they give you classes? '''Shanesy:''' Not so much; but there weren’t a whole lot of new things to learn. By today’s standards, there was nothing. I mean, the speed of change now is just stunning. In those days, there were not very many different machine organizations. There was 704, which is what I started on; and then we went to 709, 7090, 7040—but those were all the same instruction set, more or less. You know, little changes. The 1401 came in about that time; that was, I think, fallout from a government contract. That had a completely different organization, and I did learn that, because I needed to. But there wasn’t so much stuff, and the operating system was so minimal that we’d [just] get some manuals, and we’d read the code. In fact, that’s the way I learned FORTRAN. “You want to learn FORTRAN? Here’s a manual.” And that was enough! You know, it wasn’t all that complicated. But today, it’s stunning—and it’s very hard for me to keep up. It’s where I show my age. '''Abbate:''' I think it’s hard for anyone. '''Shanesy:''' I think everybody’s struggling with it. '''Abbate:''' It’s hard to keep up—or to have any memory. People who are starting out now know all the latest stuff, but they don’t know the . . . '''Shanesy:''' Actually, it’s interesting: They have a very different view of what machines do, which helps them in some regards. I think my generation is a little bit handicapped with some of the new stuff by understanding too much. [We wonder,] “Well, how is it doing that? And why is it doing that?” If you learn it in third grade or kindergarten, that never occurs to you. You learn, the way you learn language, how to manipulate it. You discover by accident, “that means this; this causes this,” and that’s okay. But I don’t come at a P.C. like that! I want it all to be logical, and it’s not—or not always. '''Abbate:''' Did it always seem logical to you, using the IBM machines? '''Shanesy:''' Yes, and I think that was one of the joys of my work. I really always enjoyed my job, and one of the things was that you got something done. You got a problem; you worked with a person that needed the results; and you produced the results. And sometimes you could actually contribute—you know, say, “Let’s try this” or, “I made up the report a different way; see if you like it”—and so you participated a little. Particularly with Gomory, I participated a lot, and it was a nice relationship. Debugging, as frustrating as it was at times: when you found [the bug], it was just such a turn-on! And we used to help each other. That was another thing: when I’d supervise—I supervised before I was actually a manager, and that was within a year of starting—I always used to tell people, “Don’t look for more than a couple of hours for a bug. Go to your office-mate and explain it to him.” Because I found that verbalizing the problem often was enough—they would stop the conversation halfway through, saying “Oh, never mind. I know what it is.” Finding a bug was great; getting the output that you were trying to get was great; more recently, when you do on-line systems or something, it’s very rewarding to bring up something and be able to play with it on-line and show it to somebody, and then to see somebody actually using it in production. I used to work with the Police Department, and you’d see these people who worked all day with the program you’d written and did good things with it, and that was great! It was a real turn-on. '''Abbate:''' Now, was it usually one person writing a program? '''Shanesy:''' In Research it certainly was. They were usually modest-sized, so the times that we would write programs together was fairly unusual, although certainly with this operating system we did that. There were at least four people involved in that: two of us heavily, and two others that I can remember—maybe even more did some work on it. We used to have this wonderful programmer, who wrote beautiful code—usually assembler language—but he didn’t like to debug, and he wasn’t very good at it. After I’d supervised him for a while, I used to just have him write the stuff, and I’d test it and debug it—because it was easy. Somehow his mind didn’t quite work that way, but he wrote the most gorgeous, compact, elegant code—better than I could write! '''Abbate:''' Well, what makes code elegant? '''Shanesy:''' What makes code elegant? Not doing unnecessary things; easy readability. Later on, people invented structured programming, which was an attempt to make code more easily read by a stranger, and so on. I never was a great fan of it; I thought it introduced some complications; but certainly, if a person was not a very good programmer, it helped a lot. But some people could just write code that was crystal clear. It had something to do with their ability to verbalize; even writing skills. In assembly language code there’s a whole chunk of the card that you write your comments in, and some people wrote really the dumbest comments, if they wrote them at all. Like “Test A”—but they wouldn’t tell you what “A” was. Whereas a better coder would say, “Test for this condition, represented by A = 1,” or something like that: write a better comment. His had nice comments, and it just flowed. '''Abbate:''' So there was an ability to articulate what you were doing? '''Shanesy:''' Oh, yes! And in fact, we used to stand around with a whip—it didn’t do a whole lot of good, sometimes, but we did really try to get people to comment programs, because it was tough if you’d got somebody else’s program to fix. I was telling you about those octal correction cards: we had a program that maintained the stock in the stock room, and it had been handed down from person to person to person, and it was regarded as a curse because it was so complex. It was assembly language, and nobody wanted to touch it; when we had to make a change, the person would go in and fix one little area, because it was so fragile. And that’s still true of code today: it gets fragile if it’s not well-maintained. You know, all your subject-matter experts leave, and you’ve got this thing that works, but heaven help you if you have to change it—or the operating system changes, and suddenly it doesn’t work! === CICS [Customer Information Control System]<br> === '''Abbate:''' Did you have anything to do with Y2K bug-fixing? '''Shanesy:''' Very little. I now work with CICS [Customer Information Control System]. Some huge percentage of the commercial transactions in the world run under CICS. '''Abbate:''' What is CICS? '''Shanesy:''' CICS is transaction-processing middleware. Transaction processing is doing things on-line. CICS is a layer that sits between the operating system and applications, and allows you to code your applications, where you’re sharing things like files, and terminals, and queues, and so on—all the things that you have to worry about when you start doing on-line processing and you don’t want two people to update the same record at the same time. You don’t want to open the file every time you want to do something; you leave the file open and under control of CICS, and it manages the accesses to it. So it’s an enabler for writing the kind of stuff we all write today. You know, when you want to pay you phone bill, or you call up the electric company and say, “My lights are off,” and they look up where it is, and what should be done about it: all of the things that you do—a lot of them—run on CICS. So it’s a piece in the middle. It comes out in versions (every once in a while there’s an upgrade), and the current versions at the time were all Y2K-ready, but some of our customers were still on older versions; so we had a lot of alerts, and we were all on call over that weekend [of January 1, 2000] in case anything bad happened. And the customer work that was done was incredible. Big companies, like the big stock brokers and so on, had to get separate systems for testing, and they tested for ages. It wasn’t really the operating system that was wrong; it was just that they had used a two-digit year. You know, who ever thought that these things would go over? '''Abbate:''' Would last that long, you mean? '''Shanesy:''' Yes. A program I wrote for the Police Department, which searched for fingerprints: there was a kind of a skeleton of it, including the file layout, done on an earlier system called “Faster”; and in those days, memory was extremely scarce—and also so was disk space—and so the designer had a four-byte date. It had four bits for the month: that took them up to twelve (in fact sixteen); they had five bits for the day; and then they had six bits left over for the year. [CS explained in subsequent correspondence that technically seven bits would have been left over, but “we avoided the sign bit.”]. [recording pauses] '''Shanesy:''' I was saying that he had left six bits for the year. You can’t be fingerprinted until you’re 18—I think it’s 18; there’s a statute about it—so he had taken advantage of that fact, and we didn’t worry about people who were older than 77 or something (we figured they were not too harmful), so six bits was enough for a number of years. One year, over New Year’s, I got this call: “Carol, the fingerprints don’t work!” And I knew immediately we had wrapped over that barrier. So that was a case of a one-[byte], or less than one-[byte], year, which was really risky, but even the two-digit technique meant that a lot of code had to be changed for Y2K. There was a huge effort—and I guess it paid off, because we had almost no calls at all. My group didn’t have any. '''Abbate:''' I just wondered if you had revisited any old code written decades before, to see what it was like. '''Shanesy:''' Well, I had an odd experience this spring. I met a fellow who had worked at the Police Department when I was there; and I hadn’t been there for 20 years. We were talking about people and so on, and at some point he said to me, “You know, we still use Beta.” (“Beta” was the name of this fingerprint program.) I couldn’t believe it! I mean, that code is actually still running and still in use. I just couldn’t believe that something that old would still be used—because fingerprint technology has advanced enormously—but it still is! '''Abbate:''' Well, let’s go back to that. This is in 1969, I think? === Working for Rand === '''Shanesy:''' In ‘69 I went to Rand. '''Abbate:''' How did that happen? '''Shanesy:''' It was a civic impulse; I had wanted to do something. Those were the days of a lot of civil unrest, and a lot of attention to poverty, and particularly urban problems. Rand, in California, which was primarily a defense contractor, had gotten a mission to work on urban problems, and they opened up a unit in New York, kind of a little branch; and when I started to explore how I might do something else that would be maybe a little more useful, Dr. Gomory knew of them. He had, I think, worked at Rand once; or he had a close friend that did, at any rate. He told me about this, and I went and got a leave of absence to work down there. I worked with the Fire Department, primarily; that was the project they assigned me to. At that time they were fooling around something called “slippery water.” I don’t know if you’ve ever heard of that. '''Abbate:''' No, I haven’t. '''Shanesy:''' It was a polymer that they put in the water that they were putting the fire out with. It had very long strings in it, and it would reduce the normal friction—which I didn’t even know about until this project—the normal friction of water, of a fluid, against the pipes. There’s a lot of friction, so that you have to have a lot of force to maintain the force at the other end—and of course you want a powerful stream, because you’ve got to get it up, in many cases, a couple of stories. So they put this polymer in the line, and that allowed them to get much higher pressure on the hoses. Unfortunately, it also made everybody fall down, on the street! [laughs.] Because it would get on the street. And the firemen were a little bit against innovation sometimes, and they didn’t really like this stuff, so I don’t think it ever took off long-term; but we did certainly experiment with it for a while. '''Abbate:''' What was the computing part? '''Shanesy:''' That didn’t have any computing element. What I worked on, first, was: The Fire Department had actually, at that time, key-punched an 80-column record for every incident (80 columns, of course, is card-sized); but they didn’t really have any programs to manipulate this data that they’d collected. We were very interested in finding out where fires were occurring and how we could deploy our resources better, based on where the fires were happening. At that time, most of the fires were reported by box: in other words, someone didn’t call; they ran down the street and pulled a box. There was a fire box on every other corner in both directions, so you only had to walk—or run—at most a block to pull an alarm. And the fires at that time tended to be in the poorer areas of the city, where there weren’t a lot of telephones. '''Abbate:''' This is all of New York City? '''Shanesy:''' This is the five boroughs of New York City, and the big problems were in the Bronx and Brooklyn and Upper Manhattan. What was happening is the Fire Department didn’t have enough resources. The way they dispatched equipment was: You pulled a box; the box rang its number—talk about crude!—it rang its own number in a central office for the borough—so Queens would have their boxes, and Manhattan would have their boxes; and a dispatcher would pull out a card for that box and radio the equipment that was on the first line of the card. That’s the origin of the term “a one-alarm fire”: you dispatched [what was] on the first line; a two-alarm fire, you went to the second line of the card—and of course, they brought equipment from further and further away as you went down to the fifth line of the card. The problem was that we didn’t have much information. When somebody pulls a box, you don’t know what it is. It could be a conflagration or a false alarm—and there were a lot of those—or it could be simply food on the stove, or something fairly minor; but you didn’t have any way of knowing, so you sent the full complement. Then a nearby box would come in, and you didn’t have anybody very close, because they were all at the first fire; they weren’t down the street. So the studies we were doing had to do with “How can we do our initial dispatch a little smarter? How can we deploy companies better?” When you go to a second alarm on a fire, companies start moving into the area; they don’t go to the fire, but they go to the empty houses. And so we were studying all that, and we needed this data to be reduced and manipulated and so on to do that. We also ran simulations. I didn’t myself work on that very much, but we used a simulator to see what would happen in various scenarios, and that resulted in a major change in dispatch policy. There are three engines and two ladders on the first line of the card, as a rule—different boxes might be different, but most of the time—so we used to always send three and two. We started, for certain boxes, sending two and one, because our experience was that was enough: the risk was low, and it was not as great as the risk of draining those houses for subsequent fires. So we changed . . . Now, of course: first of all, they get much more traffic by telephone, so they have a much better idea of what’s going on. '''Abbate:''' Right, because the box doesn’t tell you. '''Shanesy:''' Yes. But that’s what we did; we reduced [the data]. And then: the alarm assignment cards were typed, and it was a horrible process; they’d get very dog-eared, because some of the boxes would just be hopelessly busy, and they were kept in tub files. So I wrote a program to just print them, and then we keyed all the data. Then we could change them at will—you know, if we wanted to change the arrangements. Stuff like that. It doesn’t sound very exotic, but it was very, very interesting! '''Abbate:''' Well, and it sounds like it’s got an immediate practical use. '''Shanesy:''' It did; it did. I think the Fire Department, more than any other—well, I shouldn’t say that; it was the one I worked with—but they made a lot of changes as a result of Rand’s work. It wasn’t always easy, but they really did get some gain. Eventually, the last thing I did there was to dabble at the beginning of automated dispatch, in which, instead of having these boxes ring, we’d pick up the signal and put it into the computer, and then show the dispatcher what the alarm assignment code was, so that he didn’t have to count by ear and pull the card. These boxes are going off in this big room, and they ring their number four times—and three or four of them will be going off at once, sometimes—and these guys always got it on the first ring. They would call it out, and then, if nobody said anything, that was the right number; if somebody had heard something different, they would say, and they would wait for another round. '''Abbate:''' So it’s ringing the number in the central office? '''Shanesy:''' If the box number was one-four-two-five, it would go “ding, ding-ding-ding-ding, ding-ding, ding-ding-ding-ding-ding”—very fast, much faster than I’m doing it. '''Abbate:''' That’s incredible! '''Shanesy:''' It was incredible. I’d go there and I’d think, “How’d they do that?” [laughs.] But they got very good at it. They were great guys, and as a reward, when we got all the alarm assignment cards in, they asked us—this is another women’s issue: The guys in our project had all been taken out to see what was going on in the field; they’d take them out on a fire truck and so on; but women: no way! '''Abbate:''' They wouldn’t? '''Shanesy:''' Oh, absolutely not! Absolutely not. “It would upset the men in the firehouse.” Suddenly there were “insurance issues”—[as if] men aren’t insured! Anyway, when we finished this alarm assignment card, they were really very grateful; we worked very hard on it, three of us—two of us, primarily, but a third girl also. And the deputy commissioner said, “What can we do for you? We’ll buy you something; we’ll get you a camera; we’ll do this . . . “ And I said, “We want a ride!” And he said, “Oh, I can’t.” And I said, “Well, that’s all we want.” And eventually he came back and said, “I’ve arranged it.” They took us up to the Bronx to a very busy house, and the guys were very uncomfortable at first. We started at dinner—this was in the evening; it was about six o’clock—and they put a plate of food in front of me that was the biggest thing I’ve ever seen! It was a steak that was twelve inches long, I swear, and a mountain of potatoes on one of those platters. . . '''Abbate:''' Like a serving tray? '''Shanesy:''' Yes! And I’m looking at this thing, and I think, “I have to eat this, because they’re uncomfortable enough and I mustn’t make it worse by not finishing my food.” Mercifully, there was an alarm almost immediately. This particular truck had a little compartment behind the cab. Some of them ride on the gate on the back; the chief or lieutenant is in the right-hand seat, and the driver’s in the left-hand seat; but since there were three of us, they put me in the cab with them. We run down to this fire—and the Bronx is this maze of streets; it’s got all these cliffs and everything—so we just tore up and down these streets getting to this fire. It was a big fire; it was a building with four stories and apartments in the front and the back; and there was a vacant lot next to it, so you could see that it was really engaged; and there were people hanging out of windows and balconies in the front and back. So they said: “Stay here!” and they rushed in. And that just sort of loosened everything up, and we talked a lot about what it’s like to go into a fire when you can’t see, and all of the training they get. It was just a wonderful evening—and we had runs all evening. It was just busier and busier and busier, and finally one of the guys said to me, “Oh, get on the back step!” So I actually rode on the back step of a fire truck to a car fire. It was a very good period. It seemed like very useful work, and very nice people to work with, and it was great! '''Abbate:''' Who knew programming could be so adventurous? '''Shanesy:''' Yes. === Return to IBM as a Systems Engineer New York City Police Department as a Client === '''Abbate:''' Now, you also worked for the police around that time? '''Shanesy:''' I did a little work for the police at that time, but not much. When I went back to IBM I worked as a Systems Engineer, in a regular marketing office. '''Abbate:''' So they had this public sector division which was marketing to public agencies? '''Shanesy:''' Yes. In those days, in New York—since New York is a very big city—the branch offices were specialized. So there was Insurance, and there was Manufacturing, and there was Banking, and there was Public Sector. And Public Sector had, in a way, the least productive and most troublesome accounts, because city governments don’t get a lot of money. We always called ourselves “the trailing edge.” [both laugh.] '''Abbate:''' They were troublesome in the sense that they didn’t buy a lot, but they needed a lot of support? '''Shanesy:''' They needed a lot of support, and they tended to sometimes be struggling with old equipment and not highly skilled programmers, and so on. So it was sometimes a challenge working with them—particularly the city itself. We called ourselves “Med-Ed-Fed”: we were the government part, but another part of our office handled hospitals and all of the colleges and educational institutions, the School Board and so on; they were a little more normal. In that office, I spent most of my time working with the Police Department; that was my customer. '''Abbate:''' That was ‘72 to ‘83? '''Shanesy:''' Right. '''Abbate:''' You must have been there during one of the blackouts. '''Shanesy:''' I was there during two of them. Oh, it was wonderful! [laughs.] We had a fairly deaf person on our staff; he wore a hearing aid, and when he couldn’t hear, he would take it out, and start fussing with it. He was quite good at fixing it, and it was old-fashioned, and it was always acting up. [When the blackout started,] we lost the power on the machines before we lost the lights—I guess we actually had some emergency lighting—and there was Eddie: he had lost this drone that we always had from the machines, and he figured, “Well it must be my hearing aid!” So he had the thing out on the table, and he was working at it; and then I came by and I said, “Oh Eddie, it isn’t you!”—which he couldn’t hear. And then everything went. That was quite exciting. '''Abbate:''' Were you actually living in the city at that point? '''Shanesy:''' Yes, so we had a lot of excitement that day. New York was very early in doing 9-1-1 on-line. With the help of IBM Federal Systems, the Police Department had built on the Airlines Control Program to do 9-1-1 dispatching. So we had on-line dispatching, I think, first in the country. Very sophisticated system for its day. '''Abbate:''' That’s in the early ‘70s? '''Shanesy:''' Yes, it was in the early ‘70s. We dispatched, from Manhattan, all of the boroughs, and we had about 25 radio channels that we needed to dispatch to all our locations; and when we lost power like that, we had to open up the borough offices and do it the old way. Of course, there was not much power to do anything. Things were very bad! '''Abbate:''' Did they switch to telephone or some other way? '''Shanesy:''' We had telephone, and I suppose they did, but we didn’t have anything in the car except radio. I don’t remember the exact details, but the programming staff stayed all night with the cops that were on duty. Most of the programming staff was uniformed: not all of it, but most of it was; we all stayed on. I remember, about ten o’clock the next morning we drove over to New Jersey to get some food, because nothing was open and nothing was fresh. It was really interesting. '''Abbate:''' Now, what were you doing for the police? '''Shanesy:''' I was helping them do the programming projects that they wanted to do. I was a Systems Engineer, and that was right about the time that IBM was required to (or agreed to) bill for that kind of help, so I did it under contract. At least ,when I was writing their code, I was billing, but the rest of the time I was just helping them do whatever they wanted to do. We wrote the fingerprint program: you know, we converted the old Faster program and made it a robust system for searching for fingerprints; because when you arrest somebody, you have to find out who they really are—that’s the way you get the rap sheet. The State of New York also does this, but they were less accurate and it took longer because we had to fax the prints to Albany. So we couldn’t meet our 24-hour arraignment deadline unless we classified the prints and searched. So that was what our program did. '''Abbate:''' And the computer was searching a database of . . . ? '''Shanesy:''' Of all the people that had been arrested. There were legal reasons why sometimes we removed prints—if they were arrested and cleared, or the record was sealed, then you’d take the prints out—but we had a record of everybody that was active. You roll the prints, and then they’re classified—two items per finger, a count and a shape—and then you do a search based on those things. The count’s never exact, so we’d search within four of the count, and then we’d give points for the best hit. It was a very good algorithm (I didn’t invent it), and it was the first or second guy most of the time. Then you would print out ten hits—the best ten people—and they’d go to a tub file, pull the print, and say yes or no. Later I did a program to do latent prints, where you have a partial print, and you are going to wind up with a huge list of people who might be that person. That’s much trickier, because you can’t index the file; you don’t always know what finger it is, even. '''Abbate:''' So you’ve got a piece of evidence—there’s a fingerprint on this water glass, and it’s half of a finger, and you don’t know . . . '''Shanesy:''' Right. So you have very little information. Now, half of one finger, they probably wouldn’t [bother]; you’d get so many hits it wouldn’t be worth it. But sometimes you’d have enough to know that you had—maybe you’d know it was either these two fingers or those two, based on the angle. You wouldn’t know what hand it was, but you’d know that this one’s a little bit longer. And maybe you’d know that the first one was a loop, and it had a count of at least ten, and the second one was an arch. '''Abbate:''' So that the count is how many lines? '''Shanesy:''' It’s the number of lines between features; there’s features in a fingerprint. There are, depending on who you ask, either five or eight basic types, and then a count. Two of the types don’t have a count, but the rest of them have counts. And we actually got people with that program. We used to just run through the file. We had about 14,000 people, I think, on the file—which in those days was a big file! Today it’s nothing. '''Abbate:''' So, you wrote a program that would take this partial print and run through every possible match? '''Shanesy:''' Yes, right. And we would batch: we would search for all the reasonable positions within the hands, since we had to go through the file, and the idea was to bring in big blocks of records at once and do a quick check, and it was actually surprisingly efficient. If we were looking for three or four different possibilities, we’d just do them all at once, because you have the record and now you’re going to look for these ten different guys. The fingerprint technicians were just amazing; they had this photographic memory for prints. One particular case I remember: somehow, on a van, they picked up some prints, and they nailed the guy; but they had some other prints that didn’t belong to him, and one of them was an “arch.” An arch is just a thing that goes like that; you can’t count it, because there’s no feature to count from. One of the technicians said, “I’ve seen that print before”; and he went and looked up all of the known associates of this person and found him. This guy was wanted: he was a push-in rape artist, and they really wanted this guy. And just because he remembered this, and it was six months since he’d seen that other print—that just blew my mind! They were really good. So they were fun to work with. '''Abbate:''' Now, you got some kind of award for doing this? '''Shanesy:''' On another case, yes. The latent print program. They were searching for someone who lived in one of these high-rises and would follow old ladies up in the elevator and then, as they went in their apartment—they call it “push-in”—he would go behind them. He injured and abused a lot of old women, and they really wanted this guy terribly; and they got some prints, and we found him! So I got a civilian award for that, which was really nice. It was a nice day, except I felt a little guilty: most of the people getting awards were people who had fetched people out of fires, or saved drowning dogs, or something that required a lot of bravery—and this wasn’t that at all. '''Abbate:''' Instead of years of quiet toil. '''Shanesy:''' Yes. '''Abbate:''' So that sounds very satisfying, your work. '''Shanesy:''' That was one of the most fun periods of my life. I worked very hard during that time, but the guys were great; we got a lot done; it was very rewarding. Everything we did they really used and were dependent on. It was really the early days of automation of all that stuff, you know. Nowadays the Police Department wouldn’t think of not having their records in machine-readable form, but we didn’t in those days. '''Abbate:''' So you were working on a lot of different systems for public agencies. '''Shanesy:''' Yes. My main assignment was the Police Department, but I was brought in there to get CICS going. Since I was programming and dealing with it a lot, I became sort of expert in it, and so other S.E.s . . . '''Abbate:''' Other . . . ? '''Shanesy:''' Systems Engineers. Systems Engineers were the technical support to marketing. We were the techies, and the marketing people sold: it was quite a clear division. We didn’t make commission, but we were assigned to help, and when we did good, they sometimes rewarded us. Other S.E.s would call me to help them on other accounts. And we used to do that all the time: if I had, oh, maybe a communications problem, which I wasn’t very good at, I’d call another S.E. that I knew was good at it, and he’d come over. So we went around to different accounts, and I wound up getting into a lot of accounts. It was fun! '''Abbate:''' How did you end up leaving the Public Sector group? '''Shanesy:''' Well, I changed managers at some point. I guess my manager left, and I got a new one, and he decided that he should change all of the assignments. I couldn’t work at the Police Department anymore, the S.E. on the Board of Ed couldn’t do that anymore, and so on. The customers were furious—they depended on their S.E.s. There was so much upset that I just decided, “Enough!” And I’d been there for what, ten-plus, thirteen years, maybe? Anyway, a long time, and it was just time to move on. So I went on to another branch, but it was a funny branch that handled only internal accounts. IBM is one of the biggest users of its own software—I think at the time we were the third biggest user of our own software—so our people also needed systems engineering help, and that’s what we did. I worked with our dispatch system. We have a big old system: when your machine fails, you call a number and a Customer Engineer is dispatched—that’s the system that does all that dispatch work. Now the machines actually call in themselves—you know, “call home” [laughs]—and they actually report on-line data that “I’m getting too many of this kind of soft error.” So it’s a very complicated system, and I worked on it in the early days when it was mostly manual input. I worked with the Burlington Manufacturing Plant; I worked with everybody. That was interesting, too—maybe not quite as interesting, but I always enjoyed it. And it was a nice group of people to work with, all very senior. It was just a nice group! === Reflections on CICS === '''Abbate:''' Were these similar types of problems, in the sense that they were transactions systems? '''Shanesy:''' Yes. I mean, I was a CICS specialist—there were two of us in CICS, that’s how heavily it was used, and we were busy. Practically every big installation had a number of CICS systems. In those days, some of them used IMS as well. [DISC 2] '''Abbate:''' Do you think in some sense the increased importance of CICS reflects differences in the way people have used computers over time, that there’s more interactive systems? '''Shanesy:''' Well, we graduated from batch systems to on-line systems; that was an enormous increase in productivity generally. If you call up the phone company and want to know about your bill, it’s a whole lot easier for her to key something in—notice I said “her”: sexist!—and just have something come up on the screen, than to go to some room with files and start sorting through papers. That’s just one example. It was a huge increase in productivity, and companies found that if they didn’t have that, they couldn’t compete sometimes. Now, your telephone company is a monopoly, but it was [still] an enormous advantage. If you think about something like airline reservations, where you’ve got people all over the country trying to get at the same resource, it’s just the way to go to be on line, so that you can find out immediately whether a flight is available and sell it immediately. I think companies are finding the same thing with the Web now. Web service is not just a luxury for some businesses. It’s a necessity to have a good presence, because some of their customers want to do business that way. So that was all happening, and CICS was the primary vehicle for doing that. '''Abbate:''' Did the spread of networks change your work a lot? '''Shanesy:''' Yes. We used to think, when we first started with CICS, that one transaction a second was really moving! And now we talk about five hundred or a thousand transactions a second. One of the big problems in early CICS was memory, and one of the big items was the table that listed the terminals. Well, we were having trouble with five hundred terminals, and now you have networks with thousands and thousands and thousands. So, yes, it just grew and grew and grew. A great deal of commerce now is done on-line, as opposed to in the batch. The banks and some of the brokerages and so on still redo all their transactions at night, in the batch, for safety (some of them do, anyway)—but as you’re doing on-line trades, can you imagine the stock market now going back to . . . It’s hard to imagine! '''Abbate:''' Guys holding slips of paper. '''Shanesy:''' Yes. Right. It’s just a good way to do things, and so there’s been more and more of it. And the product “CICS” has evolved in the most incredible way. I mean, it’s amazing that a product that’s 30-some years old now is still the vehicle for probably half the transactions in the world that go on. '''Abbate:''' This is a single software product? '''Shanesy:''' Yes. '''Abbate:''' And it runs on all of the various types of computers? '''Shanesy:''' No, it runs on IBM mainframes. We’ve built some versions for smaller systems, but they were never very important. Its great benefit is that it allows a great many terminal users to work at the same time, and you don’t really have this problem on a PC; there’s only one person there; so it never really caught on terribly. There’s something called the “CICS client,” which is a way that your PC or practically any system can access host CICS. But it’s primarily an OS/390 product—or zOS, whatever the name is these days! [laughs.] '''Abbate:''' You started doing that back in ‘84 or so? '''Shanesy:''' Yes. I’ve made a career of CICS. And I’m still doing it. It’s almost embarrassing! [laughs]. And now I work in the center of competence for CICS, which is in Dallas, so I work remotely. '''Abbate:''' So you can live in New York and . . . '''Shanesy:''' Yes. It’s not perfect. It’s not as good as being there, but I didn’t want to move to Dallas. === Reflections on Career<br> === '''Abbate:''' Now, you spent basically your whole career at IBM, except for that stint at Rand. '''Shanesy:''' Yes. '''Abbate:''' Did you ever consider working anywhere else? '''Shanesy:''' Not seriously. IBM was, for most of my career, a very paternalistic company; they took care of their employees much better than most. That’s probably still true, but the whole environment’s different now. And I had a job that I liked, that kept me going and was always interesting. No job is fun all the time, by any stretch, but basically I liked my job, and pay was good. (I can’t even make that statement now! Things have changed a great deal.) But no, I never really very seriously considered it. Probably should have, but I’m not particularly adventurous: I never wanted to work for myself and worry about things. I just enjoyed what I was doing, and kept doing it. '''Abbate:''' Is there something you think you would have enjoyed more, that you missed? '''Shanesy:''' No, I think the thing I look back on now is: Maybe I should have worked in a different area, where maybe I could have made more difference. I mean, when I worked in Public Sector, I felt like I was doing something for the public—even if in a tangential way, but nonetheless, I did feel that way. I don’t feel that way so much now. I work for money-making concerns, as a rule—and there’s nothing wrong with that, but I think I’m just at the time of life where you look back—you don’t have very many working years left—and you say, “Gee, I wonder if I should have done that.” Probably anybody would do that. So I do have that regret—or that question in my mind—but other than that, no. '''Abbate:''' Did you ever feel that, as a woman, you might not have had equal access to pay—or training, promotion, anything like that? '''Shanesy:''' Yes, I did feel that sometimes. I must say I grew up not expecting to have equal access, and I was surprised later in life at how ingrained that was in me. My parents were probably modern for their age: my mother worked, and my father never let her [do all the housework]—if I weren’t home, she didn’t do the dishes alone; they did them together, and he didn’t have to be asked: they shared the work. There wasn’t a real fine line of “you do this and I do that.” But first of all, I’m not a competitive person by nature at all, and I just didn’t expect to get the breaks, I think. I never really tried very hard to rise up in IBM, and it was a while—I’m a little thick that way, I guess!—it was a while before I realized that you had to become a manager to get ahead. I was a manager, but once I had been at Rand and went back to what they called “the field,” I didn’t want to be a manager. I thought it was a horrible job, and it took me a while to realize that IBM didn’t offer much to people who didn’t go through the management chain—and still doesn’t, really, although they do much more than they used to do. I certainly experienced managers who, while very nice, simply seemed not to hear you. Sometimes I would suggest something and get no response. Then a man, one of my colleagues, would say the same thing, and suddenly the manager would respond. Even my male colleagues would talk to me about it—one in particular: “What’s with him? I said the same thing you did, and he jumped right on it and said, ‘Well, that’s a great idea!’” So I did experience things like that. I never felt particularly worried about it; it wasn’t a big element; but it certainly happened. '''Abbate:''' I wonder if that’s a characteristic of computing in general, that there’s not a big path for advancement if you don’t go into management. '''Shanesy:''' It probably is. I see in my accounts people that the place would fall down without, and I doubt that they get anything near the compensation that some of the managers do. You see this in brokerage houses and so on. I don’t know what they make, so I’m surmising. And certainly, even in IBM I don’t think that the very good technical people are always recognized and paid in the same way that, say, some person on fast-track management is. But I’m doing that without numbers; I don’t really know. And I don’t think IBM is any different from other companies; they’re probably better than most in that regard. '''Abbate:''' So probably not specific to IBM. '''Shanesy:''' No, I don’t think it’s specific to IBM, and I’m not sure it’s specific to computing. I certainly think girls are much more aware nowadays of what they should get, as opposed to what we were aware of. As I said: these places that only wanted to interview men? We didn’t think anything about it! Or not much, anyway. '''Abbate:''' Did you have difficulty balancing work with family responsibilities? '''Shanesy:''' Well, I married and divorced—I wasn’t married terribly long—so I didn’t have children; but one of my regrets is that I worked way too many hours for IBM. I just worked too hard. A big element of this is my personality, and the fact that I was enjoying it, and so on. So I wish I hadn’t done that. When my parents got sick, it was a strain; it was a struggle. I didn’t like to leave my mother. My father died first, and then my mother had Alzheimer’s, and she was ill a very long time—thirteen, fourteen years—and it was hard for me to travel; and by and large IBM was very good about it. I would travel, but I wouldn’t go for very many days, and they understood that. I think it was just hard on me, period; I don’t think it had much to do with work—although I was working very hard also—but they wouldn’t have forced me to work that hard. [recording pauses] '''Abbate:''' We were talking about women struggling with work and family. '''Shanesy:''' Especially when the kids were little. That was the hardest period for the women. And that did land more on the women than the men—just the nature of our society. '''Abbate:''' What kind of hours were you working? '''Shanesy:''' You mean in terms of how long? Ten or twelve hours, often. '''Abbate:''' A day. '''Shanesy:''' Yes. And particularly at the Police Department, [because] I used to like to hit two shifts. The way they worked was, they’d work a week on eight to four, and then the next week would be four to twelve, and then the third week would be twelve to eight; and so if we were working on a project—I didn’t usually try to hit the midnight shift, but I would come in at ten, and then work till maybe eight or ten at night, so I’d catch both shifts, and we could work on projects that way. That was just a quirk of that particular account. Then I’d get a ride home, at least! '''Abbate:''' In a police car? '''Shanesy:''' No, just in a policeman’s car. '''Abbate:''' Do you think computer jobs have become more open to women since you started? Or less open? '''Shanesy:''' I think they’ve become more open. I certainly see many more women that I used to. And they seem to be—I think part of this is effort on companies’ part—but they seem to be in management positions. It’s not unusual at all to find a 50-50 ratio. Now, different businesses are a little different in that regard, and I often still find myself in a meeting where I’m the only woman; but it’s much, much better than it used to be. So I think things have opened up considerably. === Reflections on the Field of Computing === '''Abbate:''' How do you think the field of computing in general has changed since you started? '''Shanesy:''' Oh, wow . . . Well, it’s become so pervasive, and so complex, in the sense of the number of products and the number of possibilities of how you do things. It used to be, if someone had a way to do it, then you did it that way, and you felt very fortunate; but that’s really not true now. You have loads and loads of options about how to implement a particular application, and part of the difficulty is choosing the best one: guessing what the future will be, and where you should invest, and how much effort you should put in, and what kind of equipment you should buy; it’s gotten very complicated. And people make terrible mistakes—but we don’t know what’s going to happen in the future either. At least I never steer a customer toward a solution I think isn’t going to be long-lasting, but sometimes something comes out with a big splash, and we think it’s the answer to everything, and guess what? It’s gone in a year! I’ll give you an example: C++, C language, I mean; I could give you a whole discussion about that, but that was very big for a little while, and almost disappeared; it’s still there, but most use has evolved into Java. Now Java is the current big thing, and that one thing, I think, is going to last; but many, many similar things came along, and were taught in the universities—that’s often where they’re born—and then disappeared. We’re still coding in COBOL, for heaven sakes! The oldest, clutziest language! But there’s this huge base of people out there that know it; it’s a very good language, actually, for a lot of things. I’ve drifted away from your question, but: There’s just so much out there now, so it’s a very, very complex arena. And lots of specialization; very hard to know all of what’s available. '''Abbate:''' What have been the most satisfying aspects of computers for you? Of working in the field? '''Shanesy:''' I think . . . Every time I’m on a project where we do something that really needs to be done, that’s a big turn-on. Finding a persistent bug is a turn-on. Helping a customer. One of the things I do now—not very often, but sometimes—is: some customer will call up, maybe a broker that’s got this big on-line system. When one of those brokerage systems fails, it’s very expensive for them, because they have to sell at the price in force at the time, and if they have to buy something, they have to buy it at a higher price; they can’t charge the customer, they have to eat that. And it gets in the newspapers, and they all get very excited when these systems go down—as they do. And sometimes we get dispatched, and if you can help the customer find the problem and fix it, it’s a big turn-on, because it was very important to them, and so you did something good. I think that’s one of the nice things about the job. The other thing is, I’ve worked with a lot of people that I’m still friends with, and that I just might not have met in another job. I still know a lot of the cops that I worked with (they’re all retired and on to other jobs now), and I know people from my old branches, and I’ve mostly worked with very nice people. That’s a big plus in a job! '''Abbate:''' It sounds like programming has been very social for you, which is not the reputation it has. '''Shanesy:''' No, it isn’t; but when you work with customers, there is a lot of social interaction. And in fact, especially when you get called into a fire-fighting situation—you know, where somebody’s down, and they call you in as an expert to solve a problem that the local staff can’t solve—that’s not the best introduction, because the local staff resents it. They’re embarrassed that the cavalry had to be called in. And so you have to interact, and you have to be careful not to make them look bad, just to help them get over the problem and, if it’s necessary, explain to them how they got into that problem. You know, things like that. So there is a lot of social interaction that you have to be aware of. Marketing people, of course, have a big element of that. Techies don’t have as much, but when you work with customers, you have some. And in the case of the police, at any rate—I was there all the time; I didn’t go to the office, ever; so that’s your work life. And we did things socially, too. I knew all their kids, and so on. '''Abbate:''' Do you have any advice for young women contemplating a career in computers today? '''Shanesy:''' Oh, golly! Nothing very profound. My advice would be not to make some of the mistakes I made. Don’t work ten hours a day. Get that balance early. I finally caught on to it, but—get that early. And request your rights. These girls are much more aware than we were; I think they come from a whole different place than where we did, so it’s not so much of a problem; but they do need to be aware that there are still difficulties, that men feel threatened by women sometimes, and there can be all kinds of subtle problems. So I guess just be aware. I don’t have any really profound advice . . . And keep up technically! That’s the other thing. '''Abbate:''' Would you recommend computing as a field for women? '''Shanesy:''' Oh, sure! I would recommend it to anybody who has that bent—who likes programming, and problem-solving, and that kind of thing. Yes, I would recommend it. And if you want to get ahead, then be aware of how people do get ahead. If it’s going into management, then grit your teeth and do it, even if you don’t want to do it. Try it for a while; if you don’t like it, then you get out. Everybody has to decide: “Do I want to make money? Do I want to make a lot of money? Do I want to make just enough money to be happy and comfortable? How do I want to spend my free time? What are my family’s needs? How important is it to me to be very important in a company?” All those things. But their mothers can tell them that! [laughs.] '''Abbate:''' Let’s hope so! '''Shanesy:''' Fran [Allen] will give you—I mean, Fran is a good mentor, and will give you a lot of insights when you talk to her. '''Abbate:''' Well, that’s a good question: Did you have anyone, men or women, that you considered mentors? '''Shanesy:''' No. I never had a mentor. I think that today somebody would have been assigned, at least for a while. You know, it depends on how you come into the company, but if they think you have potential, they would probably do something along those lines. But in my day I don’t think it even occurred to them with men, and it certainly didn’t occur to them with women. They really did expect us to get pregnant and leave pretty quick! '''Abbate:''' Really! That was IBM’s attitude? '''Shanesy:''' Oh, yes. I think it was. I mean, I didn’t hear it explicitly stated, but we always felt—and I include myself—that the men were likely to be around longer than the women. It’s not even actually true anymore, but it was then. '''Abbate:''' So did most of the women just stay a few years and then leave? '''Shanesy:''' A lot of them that I was hired with did. Now, a lot of them later came back, but rather much later. Not all. Some of the women I worked with came in as forty-year-olds—they’d done some work at some point in their lives, they’d had a family, and then the kids were grown up and they decided they wanted to work again. I worked with a number of women, when I was in my twenties, who were in that category, and they were great ladies. '''Abbate:''' Interesting, so: So they were able to take time out to have kids and then come back. '''Shanesy:''' Right. But of course, at the time they started, computing was not even an option—wasn’t there. '''Abbate:''' Oh, these weren’t people who had been in computing and then came back? '''Shanesy:''' No. Because when I started, there wasn’t a whole lot—because I started in ‘59. So these were people who had probably been out of the work force for maybe ten years. So, they were doing something before that. I know one of them was a teacher; one of them was a chemist; I don’t think I know what the others were. So they had just come in, after their families grew up. We had two absolutely amazing key-punch operators that sort of ran the whole place. We called them “The Two Margarets”: they had the same first name and they looked alike! [laughs.] Unrelated. And they came in when they were probably fifty, and they were one of our mainstays. '''Abbate:''' I hadn’t realized that. So, IBM would recruit and then hire these older women, or they got these jobs somehow? '''Shanesy:''' Well, I think, in many cases they were local women, and they’d apply for a job. IBM didn’t actively recruit them, I don’t think, but they were hiring; IBM was doing a lot of hiring in those days. They would recruit at college campuses, which I think they still do, but there were lots of other ways that people came into the business—including from customers, occasionally, although at Research we didn’t get much of that. But in the field, customers often wanted to come over—and IBM has to be very careful not to steal; I mean, they really take great pains about that, and they won’t hire somebody unless there’s mutual agreement that it’s okay. But these were just people who were local. In a number of cases their husbands worked for IBM, so they knew what was available, and they came in. '''Abbate:''' Interesting. It must have been something to be a 50-year-old computer operator in 1959! '''Shanesy:''' Yes! But they did. Now, the people around the big machine were always men; they didn’t have . . . I mean, I remember when the first woman was allowed in the machine room to really run the stuff, because that was considered men’s work. The key-punch operators were women. '''Abbate:''' Wow. And when was that, more or less? '''Shanesy:''' Well, it was still when I was at Research, so it was probably 1967 or so. There was one lady: she was a ball of fire, and she announced that she wanted to hang tapes and do all those things—which of course anyone could do! I mean, there’s no great trick to it; it wasn’t heavy lifting or anything. But it was just—I don’t know—there was this odd distribution of duties. '''Abbate:''' In terms of the space inside the building: You had machine rooms—I don’t know if they were in the basement or where they were, but you’ve got these machine rooms, and there’s just men in there actually operating . . . '''Shanesy:''' Yes. Pressing the buttons. '''Abbate:''' Did you have a room full of key-punch operators, and they were all women? '''Shanesy:''' Well, no. In Research we didn’t have a great many. We might have three—two or three—and they were always women. And then there was a fellow that was there for years and years who was handicapped; he’d had terrible polio, and so he just couldn’t be a machine operator. He walked, but it was with extreme difficulty—so he used to wire the boards and stuff like that, and I think he supervised the key-punch operators. '''Abbate:''' And then you had the programmers. '''Shanesy:''' Programmers just had offices nearby—wherever was handy. '''Abbate:''' And that was a mix of men and women. '''Shanesy:''' Yes. It was more men, and actually it was more men as time wore on, but I think that was just accidental, pretty much. It was probably a quarter women and three-quarters men, or a third women. '''Abbate:''' And then you had the scientists or mathematicians, who were pretty much all men? '''Shanesy:''' Largely men, yes, at that time. I don’t think that’s true anymore. I know there’s a woman that’s head of the math department now. But at that time, lady mathematicians were few and far between—and also physicists; that just wasn’t an area that we were getting female job candidates for. I think the people that IBM hired were pretty top-notch people, and they weren’t hiring women—whether it was from lack of candidates, or viewpoint, I have no idea. Probably a little of both. '''Abbate:''' There couldn’t have been that many women with Ph.D.s in math. '''Shanesy:''' No. I think that there were very few, in fact. ‘‘‘Abbate:’’’ All right. Well, I think I’ve probably asked you all the questions that I should put you through . . . '''Shanesy:''' Oh, it was a pleasure, Janet! It really was. I don’t know if I’ve given you anything of use, but it’s interesting to hear your questions. [END OF RECORDING] [[Category:Profession|Shanesy]] [[Category:Computing and electronics|Shanesy]] [[Category:Engineering and society|Shanesy]] [[Category:People and organizations|Shanesy]] Return to Oral-History:Carol Shanesy. Retrieved from "http://ethw.org/Oral-History:Carol_Shanesy"