Return to main page

Draft - Feb 23, 2017

1401 Restoration Oral History Panel


Bill Flora, George Ahearn, Frank King, Allen Palmer,

Ed Thelen, and Ignacio Menendez

Interviewed by:

Robert Garner

Recorded: July 1, 2015

Mountain View, California

CHM Reference number: X7542.2016

© 2015 Computer History Museum

1041 Restoration Oral History Panel

Robert Garner: All right, let's start. OK, it's Wednesday, July 1, 2015. And before me are six representatives of the most wonderful group of people I've worked with in my professional career outside of being paid. [LAUGHS] I've known everyone here on the panel for up to 10 years or more except for a couple of new people on the side.

Sitting here before me we've got Bill Flora, George Ahearn, Frank King, Allen Palmer, Ed Thelen, and Iggy Menendez. And the way I'm going to format this today, the way I'd like to format this today, is that the first half of our interview session we're going to talk about, for each of you, talk about when you joined IBM, about what you kind of did during those early days, and then maybe one or two stories. And since there are six of us and any one of us could tell us stories for three hours, maybe we'll just choose one or two. And then we'll take a short break after that.

And the second half, we'll talk about the 1401 restoration project. What was it like when you first heard about it? What was it like bringing out the machine, a couple of stories about restoring the machine. You know, Allen wrapped in tapes, Frank with his head inside the hydraulic unit, Bill inside-- we'll save that for the second half, I think. So I'd like the first half to be more about your stories in IBM: When you worked, what you did, what it was like having to run an entire business and being responsible for keeping that business up and running, that kind of stuff.

So I thought we would go sequentially. If you want to ask anyone a question, though, go ahead and ask them. So don't act like only that person can talk. If it spurs you to ask them a question about their experience, go ahead and do it. So I'd like to start with George. George. Again, they would like you to repeat the question. So you'll say who you are.

George Ahearn: I'm George Ahearn. And before I joined IBM, I was in the Navy assigned to an engineering project at the National Security Agency.

Garner: What year was that about?

Ahearn: I got there in 1956 and my tour was up in 1958. And the difference is significant, in this respect. In 1956, there was a lot of opportunity for engineering jobs. A lot of people were hiring, including the agency. In 1958, it all turned around and went negative. There were hardly any jobs available.

And that put me in an unusual position where I had to get a job. I didn't want to stay at the agency. I didn't want to stay in the Navy. But I became very impressed with working on IBM equipment at the agency. So I tried real hard to get a job at IBM. And I was striking out.

Garner: About how old were you?

Ahearn: Pardon me?

Garner: About how old were you at that time?

Ahearn: 22. So as luck would have it, I had a colleague, Holger Jensen, who was in the Air Force under same circumstances as I. That is, he had been doing engineering at the agency and he was ready to leave the Air Force and return to IBM, where he had been for a few months before he went to the agency. And he was going to be returned to a very important project in Endicott. In fact, the top project in Endicott. Which was a contract with the agency.

Garner: The NSA?

Ahearn: National Security Agency. And the man who was in charge of that project was Bob Evans, pretty high-level manager, very important, powerful manager. Holger went to Bob Evans and said, "Look, I have a friend here who's really trying hard to get into IBM and he's striking out. Do you have any interest?" And he said, "Sure, right. Let's get him to an interview."

And so they set it up. Because Bob Evans and a couple colleagues were visiting the agency on some business matters. And they felt it would be convenient to have an interview at the Silver Spring Hotel before they returned to Endicott. And they set it up for, let's say, a Wednesday of a week.

And I had to make a choice. I knew all about IBM and how you use the white shirt and the tie and the suit you look your best when you went for an interview. So I took the only suit I had and I put it in the cleaners. And that left me with sport coat and slacks and a Navy uniform.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 2 of 38

1041 Restoration Oral History Panel

The next day I get a call from Holger and he says, "Look something came up. These gentleman have to return to Endicott and so we'd like to advance your interview from Wednesday to Tuesday." And I said, "Oh, Holger, I'm in trouble. My suit's in the cleaners. I can't get it out. And I sure don't want to show up at an IBM interview in sport coat and slacks."

And he said, "Well, here's your choice." He said, "If you don't want to do that, then you're going to have to get on an airplane and schedule the whole thing and go to Endicott and have an interview up there. But that's what it will take for you to get a job." I said, "OK, I give in, I'll do it." So I made my way over to the Silver Spring Motel. I knocked on the door and Bob Evans answered the door in his underwear!


Peter Fagg, one of his subordinates, was laying on the bed without a tie on and Jack Nordly, the third man, was sitting in a chair. And we had the interview and it went very well. And what they explained to me was that IBM has a policy that no one of us can hire someone without a second look by someone else. So I did then have to go to Endicott by airplane.

And I wound up with an interview in Owego, which was the federal systems division, as well as the one in Endicott itself on the project that was being put together for the government. It was a new project. They were hiring. They had a lot of priority in the lab. And it worked out. I was very glad to go there and become part of that project.

Now, started out myself and Paul Farbanish and a few others were assigned to-- the manager's name was Wayne Brook. He was a project engineer. And in the hierarchy of things at IBM in those days, a project engineer was a first level manager. And his boss was this development engineer and his boss was a senior engineer and so on.

By the way, Evans had a thing about organization. He didn't like a deep organization with a lot of management. And he admired the Vatican. He said, "What I like is there's the pope, there's the cardinals, the bishops, and the priests." He said we should have an organization like that. Trim. Anyway, we were assigned to the IO attachment to this government project.

Garner: This was Maidenform?

Ahearn: Well, I'm going to tell you about that.

Garner: Oh, it's not that one.

Ahearn: I'm going to tell you about that. Robert's asking me to tell the name of the project. And I'd like to explain it this way. In code breaking, there's a condition called a bust. And what that means is a mistake, usually a human error.

For example, if the operator is handed a message and he's supposed to send it out, if he starts out sending it out in clear text, just as it reads, and he catches himself and says, oh, I gotta go back and do this right. If he returns to it and sends it out encrypted, that becomes a bust. Because the code breakers love that. So I would ask my fellow colleagues here, what might be a good name for a project that involves a bust?


Do you give up?

Allen Palmer: Yes, yes.

Ahearn: The project name was Maidenform.


Anyway, Paul Farbanish and I were dispatched by Wayne Brook to go around the whole Endicott lab and look for what was going on in the area of input/output. And we found out there was a project that was developing a chain printer. We found out there was a project developing a card reader punch. And we found that out there were other things like bank sorter readers for reading checks. So we went around and looked at this.

And then we came upon a small project called SPACEPACE, which was putting together a system that we now know as the 1401. And it had the IO devices that were necessary. That is, it had a printer and it had card input and output. And that's all they were interested in for the attachment to this Maidenform project.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 3 of 38

1041 Restoration Oral History Panel

So we returned back to Wayne Brook and told him that it looks like a smart thing might be to just pick up that project and develop an interface between it and Maidenform. And then you would automatically have your IO in one package. And they brought that to Bob Evans. And he was delighted with the idea. It helped him in his schedule, it helped him in his effort and so forth.

And he went down to see Chuck Branscomb, who was the area manager in charge of the 1401 development. And he said, "Chuck, what do you need? I like your project, I want it to succeed. I need it. What do you need me to do?" And Branscomb said, "I need engineers." So Paul Farbanish and I were told, "You're on temporary duty. Wayne Brook is your regular manager, but you're down there taking direction from the management on that project." All this was taking place in August of 1958.

And as things developed, in December of 1958 the National Security Agency canceled the project. I don't know why. I can speculate. But it all ended. So Paul Farbanish and I became permanent party on the 1401 project. Other people were dispatched in other places. Like there was the 7070 going on. And a few of them got sent down to the 1401 to do things like adding magnetic tape and doing multiply, divide, and future efforts. And that's how I got there.

Garner: And you were responsible for what part of the 1401 logic?

Ahearn: I reported to Russ Rowley, who was the manager of the central processing effort. And we got involved in registers and cycle controls and arithmetic and logic unit. I didn't have anything to do with the edit command. One individual was doing that. And he was getting his direction from Fran Underwood.

Fran wasn't a manager, but Fran was the man who had all the ideas about the architecture of the 1401. So the edit was really a single individual with guidance from Fran. The other thing that was a single individual was Tom Cooper was doing the memory and the clock. And if you wanted interface with the memory or the clock, you dealt with Tom. And he did that all by himself.

Garner: What kind of person was Fran Underwood like? Do you have any one story you could remember of Fran?

Ahearn: Yeah. Well, Fran was a very brilliant engineer technical person. He was not a manager at that time. He had the ideas about how this thing was going to go and what should happen. So if you had a question, you'd go to Fran and ask him that or you'd go to Russ Rowley and he'd go to Fran. So Fran was very much involved in how the whole thing went. And then the managers like Russ Rowley would get it implemented. nnnnnnnnn

Garner: Is there one Fran Underwood story you might tell? mmmmmmmmmmmm

bbbbbbbbbAhearn : I considered Fran a busybody. It annoyed me sometimes when he'd come and look over my shoulder and critique my design. I mean, he was right. He was very good at logic design and he taught it in the engineering training program. But it just kind of annoyed me at the way he went about it. I guess what I was thinking of is, you're not my manager, why are you bugging me?

Garner: He had his finger in everything.

Ahearn: Later on when the 1401 was announced, Fran became a very valuable commodity with a lot of regard for him. And they went to Fran, I suppose, and they said, "Fran, what would you like next?" And I think he said, "I want to be a manager." And that's too bad.

Garner: So what did you do after the 1401 program?

Ahearn: I moved on to the 1440 program in about the same role that I was in the 1401. And then following that, I became a manager and worked on an optical character recognition project called the 1285, which was reading journal role tapes optically.

Garner: Before that, did you work on the IBM 360 model 30?

Ahearn: Oh, yes. What happened was before the 360 was announced, and it was several years or certainly months later than the 1401, it became obvious that there was a pretty large investment in 1401 software on the part of customers.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 4 of 38

1041 Restoration Oral History Panel

And there was great concern about how the 360 could make the transition for customers from the 1401 to the 360. And there was competition from, I believe it might have been Honeywell. That they had a product that would make the transition easier than the IBM plan would allow.

So the people decided that there ought to be an emulator for the 1401 running on at least the 360 model 30. And I was assigned to be part of a team that developed that emulator. I think there were five people involved. And my role in it was to interpret to the other four people how the 1401 worked. What each in the instructions meant and how they worked. And the others were doing the micro coding in the 360 to emulate those instructions.

Garner: OK, and then by the 1970s, did you make a transition coming from a CE to a software CE? And when did you retire from IBM?

Ahearn: Oh, well, in the 1970s I was-- well, I should back up and say, how did I get to San Jose? There was a project, a 1401 project, under Fran Underwood for the attachment of a 1405 disk drive to the 1401.

And I was assigned to work on that project with several others in Endicott, namely Jud McCarthy and John Young and a few others. And then a team from San Jose made up of Bill Benz,z John Hammel, Jerry Waldorf, and Malcolm White. They came from San Jose and lived in Endicott with us for about six months. And they merged.

The two teams that became one really and developed the attachment of the 1405 to the 1401. And when it came to a certain point, management said we want to transfer the whole thing back to San Jose. And they came to me and said, and we want you to go with it.

And I wasn't too thrilled with the idea. We had just built a new house. We had moved into it in Crestview Heights. And I really wanted to live in that house at least for a while. And at that point, the San Jose people told Bob Evans that, hey, your guy back there, Ahearn, doesn't want to go.

And I was in the doghouse with Evans. Because what Evans wanted, Evans got. And if you got in his way, you're in the doghouse. So I got away with it. I still had a job and I did the things I described, like 1440 and optical character recognition.

But about six years later, the house wasn't big enough. And I got tired of the snow. And in the March time frame, it got dirty and sloppy and I wanted out. So I called up the manager in San Jose, [INAUDIBLE] was his name, I said, now I want to come. And he found a job for me. In the disk business. That's kind of my story.

Garner: What year did you retire from IBM?

Ahearn: I left IBM in 1984 in a window. And that was the year that they said, "Look, we're cutting back, and we welcome anyone who's got 25 years in to take advantage of this window and we'll give you two years salary spread over four years." And I jumped on it. And they said, oh, by the way, also, if you have 30 years, you can retire and get those benefits. Well, I didn't, but I jumped on the other thing I went with it.

Garner: And then you eventually moved to Menlo Park. Did you always live in--

Ahearn: I never worked in Menlo Park.

Garner: Los Gatos. You moved to Los Gatos.

Ahearn: Oh, I passed through Los Gatos with a desk. Sat there [INAUDIBLE].

Garner: All right. Thank you, George. That's, by the way, for everyone, that was a perfect 16 minutes to give you the sense. Frank, do you want to go next? How did you get involved with IBM? What's your story?

Frank King: Well, I got called into the army from a job I had.

Garner: About what year was that?

King: That it was in '57.

Garner: How old were you again?

King: I was 21 or two. 21.

Garner: OK, similar to George.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 5 of 38

1041 Restoration Oral History Panel

King: And was planning on getting married, live a normal life.

Garner: Married so young?

King: Yeah. And then I got drafted into the army. So then I wound up getting involved in crypto in the military. And that was a real good move because it was--

Garner: Speak up just a little bit.

King: It was about a year of school spread over about 14 or 15 months. And so when I got out of the military, I interviewed around and--

Garner: Was it crypto on IBM unit record equipment?

King: No. It was pure military crypto. No, we had class A and B. And then I stayed there and taught because I didn't have enough time left in the army to go anywhere. So I taught crypto and then got involved in a couple of projects of telephone crypto. And so after that, when I got out, I went to work for IBM.

Garner: So what year was that?

King: Several companies--

Garner: What year?

King: 1959.

Garner: '59, OK. That was the same year the 1401 was announced. Do you recall its announcement at all? Probably not.

King: No, I don't. I'd heard that IBM was a pain in the ass because of the way you had to dress.

Garner: White shirt and tie.

King: Well, it wasn't that. It was that they controlled your evening wear also.

Garner: You couldn't drink, you mean?

King: No, you could drink. You just had to dress to go out in the evenings. And I found out that was just a marketing rule. That wasn't a field engineering rule.

Garner: So they wanted IBM people to look their best even if they weren't on the job at all times.

King: Right. When you interface with a customer or run into a customer or anything, when you stayed in a hotel, you had a white shirt on for dinner. Anyway, once I got past that, it was kind of bidding between Western Electric and IBM and finally decided on IBM. Went off to unit records school into their basic.

Garner: Where was that? Endicott?

King: Endicott, New York. That was where everybody on the east coast went to basic school. You get up in the morning and you sing songs.

Garner: The famous picture of the Endicott building there, the education building, the insignia on the steps. Every Saturday morning you sang the IBM song?

King: No, every morning.

Garner: Every morning.

King: You went down and sung IBM songs for about it took up about an hour of the beginning of the day.

Garner: How did you feel after an hour of singing? Tired or ready to go?

King: No, it was refreshing. It was just a tradition. And everybody there went along with it.

Garner: So you felt like a family in some way.

King: And you felt like you were part of something. But anyway, after I finished basic, I went back to a branch and just--

Garner: Which branch was that?

King: Greensboro, North Carolina. And was working on unit record equipment.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 6 of 38

1041 Restoration Oral History Panel

Garner: Was that near where you grew up?

King: Yeah. It was about 20 miles from my hometown. And working on unit record equipment. And I worked in that for about two years. And all of a sudden, the Kennedy situation came along where he was so-- they needed us back in there. Anybody who knew anything about making crypto work was back in the army.

So I spent another year back in the military. It was an open ended. And then they let us out in about a year. I came back to IBM and I talked to my boss and I said, look, if I'm going to go to any schools for a long period time, rather than renting an apartment here, I can get the per diem and not have an apartment. So they sent me off to the longest school they had, which was right then, for that little town, was 1401 school. And that's how I got involved in 1401.

Garner: Again, this was in Charlotte?

King: Yeah, but I was a year or so behind. 1401 had been out for a while and had been shipped in the office. So most of the experience on the 1401 was with people other than me. So I wound up with a couple accounts that had a 1401 and did preventative service on it. But the failures were rare.

So I got not very much experience on the 1401. But the printer took a real careful mechanical touch. And I guess I had that at the time. And so I got any time somebody had problems with a printer printing straight, they'd call me and I'd go make it print straight. And so I got a good reputation about a printer.

Garner: Anywhere in the southeast?

King: Well no, mostly in the Greensboro area and Burlington and Charlotte.

Garner: Did you find yourself doing that every day?

King: No. It was a couple of days to fix a printer sometimes to make it print straight. And then you got to working with all the people. When the faster printers came out, the customers wanted to print eight carbons, like the 1403 model two would. And the high speed printers wouldn't do that.

So I got involved with a lot of corporate people who made paper trying to make it print the right way. Then I just continued to work on miscellaneous things with IBM during that period of time.

Garner: I had an impression of talking to you over the years, Frank, that there were a lot of calls you had to take late at night or in the early mornings sometimes?

King: Yes. Because Greensboro wasn't a big office. It had one shift. But customers were in 24 hours a day. So there were a tremendous amount of calls at night.

Garner: Like at 2:00 AM or 3:00 AM?

King: 3:00 AM. Whatever time it was, you took your turn on that. And some of the problems with the 1401 were-- a major one was when everybody cut their tapes 0.498” wide. Everybody did. This was because a magnetic tape is 0.500……

Garner: Talking about the width now. So they're nominally half an inch wide. 480 over 1,000.

King: It's 0.500 minus two plus nine. So everybody cut those things at 0.498.

Garner: You mean after they purchased them they made them narrower?

King: No. When they manufactured tapes, every company that manufactured tapes cut them 0.498 except for this company that IBM helped establish when they had helped companies get started to compete against them.

Garner: The courts had told IBM to do this?

King: Well, this was back before I went to work for IBM. They agreed to put people in the card business and the tape business and making control panels business, plug boards. And this company real close to Greensboro started manufacturing magnetic tapes. And they decided since everybody could cut them exactly, I mean, they decided to cut their tapes half inch. That's what they [INAUDIBLE]. Of course that was 0.0022 of an inch wider than everybody else's.

And one of the things that happened, of course, is the IBM heads had this big slot groove where they'd been wearing over the years. And these tapes wouldn't run because they would curl up on the edges

CHM Ref: X7542.2016               2015 Computer History Museum               Page 7 of 38

1041 Restoration Oral History Panel

because they were too wide. And that was a really major problem, much more major than I thought it was at the time.

Everybody knew back (Blackwatch?) tapes were bad. Well when I figured out finally what was happening, I realized that IBM was in a big problem. Because the company could never get a government contract because they would submit 12 of their tapes and of course they would test them on IBM machines with big slots worn in them.

Garner: And then later on new machines, they wouldn't work.

King: Oh, new machines they would work beautifully. But only the tape drives that were worn out, which most of them were. So anyway, I made a few trips back and forth trying to explain that all the way up to Endicott and to high level people that you're in trouble. You've almost bankrupted this company because they built their tapes correctly and you've been building them wrong. Anyway, that was a big, big deal for me. I left Greensboro in '74 and came to San Jose as a service planning rep and to work on a problem with software.

Garner: Did you transition to working out 360s then while you were in Charlotte?

King: Yes, I went from 1401 to 360. I taught 360 for a while.

Garner: What were some of the interesting 360 challenges? I know that the model-- one of them had the capacitive cards, the model 30. And there were a lot of issues with static electricity.

King: Well, static electricity was a problem. But IBM. Well, it was defined well what the problem was and what the floors had to be like to stop producing static electricity.

Garner: Conductive floors.

King: One of the major companies that was my customer designed a carpet that was for raised floor. The two foot panels. And with the exact amount of resistance that IBM needed. And that put just about all of the static electricity's in the town I was in to bed. Because they plastered that place with carpeted floors, which sounded bad to start with because the carpet was what was causing so many static problems. And they had fixed the problem. But one of the problems we had, I remember one problem in particular that we had that there was a lady who worked on--

Garner: A computer operator.

King: A computer operator. She worked in a large computer center, which was Burlington Industries, which was the biggest company we had. And they had three computers.

Garner: Were they 360 models?

King: Model 50, model 65, and a model 30 for slave work.

Garner: So low end, medium, and pretty high end.

King: Right. And all of them were hooked together with emergency power. But they were all really somehow connected through. And this lady had characteristics that she would build static electricity. And everybody knew, not said out loud, that when she was working was the only time you would get these phantom computer checks. And it was all over the room and--

Garner: So people associated her being in the room with computer failure.

King: When she was on shift, you would get off the model 65 or disk drive, one of the control units would blow up or you'd get a channel error on another computer. But she had to be working. So anyway, my bosses decided that I should help fix that problem, because they'd had a lot of meetings. They didn't want to fire this lady. But obviously you can't keep her working if these computers keep blowing up.

Garner: Yeah, why would she have that magic touch?

King: And so I ordered some equipment from Region, which is a static electricity meter that you could measure someone's static charge or a static charge on any device. And we had to do this kind of secret, except she had to know about it, I had to know about it, and there had to be a couple of operators that were sworn to secrecy. So we went out.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 8 of 38

1041 Restoration Oral History Panel

And I was told that you can't embarrass this lady. And I had met her before and stuff but never had a real relationship with her. She was a nice looking lady and well built from a female standpoint. And so when they got all ready, had all this equipment to see what was going on, she was wearing a white dress, obviously no underwear. And she announced it to me. And I just went bonkers. I didn't know want to say or anything. She said, "I don't have on any underwear and this dress is pure cotton." And she was barefooted.

Garner: You turned the color of your [bright red] shirt right now.

King: That's right. I had all this equipment to measure her body, how it created static electricity. And I put her on a piece of plastic. And she would build a charge. And you had this device in front of this meter. You had to keep it exactly 12 inches from her to get the measurement. And this thing would slowly start climbing and would go all the way until the meter [INAUDIBLE].

And so anyway, what had really happened is this lady, because she was so good looking, is the operators would come over and flirt with her, the operators from other machines, and the piece of plastic that she had on the floor for her chair to roll around on was not grounded. Just oh, it was even worse than I said. It was only when she was working on CPU A. If she was working on B, you never had these crazy problems. But I didn't put that together with anything, because it wasn't that much.

Anyway, if you stood on the same chair, same plastic she was on, you would build a charge yourself. Slowly. And you could walk across the room, probably from here to the wall over there, touch a disk drive, and it would blow the channel right out. Because you had built that same charge that she had and you didn't discharge it because she always was barefooted. So you didn't discharge it as fast.

So anyway, it was just a corner of the table wasn't sitting on the plastic. We moved the plastic underneath the corner of the table and that discharged the surface, therefore discharging her charge. And I've heard since then many people who built static electricity on their own. But that was my, I guess, my most famous trouble because you got so many executives both with IBM and with her around watching what I was doing. There's a few others I could talk about, but they're not--

Garner: So then you went on to 370s. And then I think IBM made a transition to transition you to a software CE?

King: Yes. Well, it was part of something called EFECEFEC, which was an extended function customer engineer. So you worked on software as well as hardware for the customer. But I went off to software school early, OS release one.

Garner: So this is the late '60s?

King: 360 OS.

Garner: The late 1960s when you went off to OS/360?

King: Yes. Whenever the first operating system was released. And it was horrible. It was not only horrible, but the problems. And because I was both hardware and software, you didn't work on software for a while and then you would go off to be an expert on it. And as you got promoted, you just couldn't keep up with all of the things that IBM had to do. So I just rolled over to teleprocessing and just made that my thing. And that's what I spent most of my time on.

Garner: Modems, special units that were meant for communication over phone lines between sites?

King: Yes. That was the most interesting problem. Because you had a lot of competitive equipment, different brands of computers trying to talk to each other. And some of the problems were just fascinating.

Garner: I think you told me about a case where it would drop random data or something and the companies couldn't figure out what was wrong.

King: Well, Burlington Industries computers communicated with the JCPenney and General Motors. And nobody really ordered their sheets and pillowcases. So the computers did that. When you would go into JCPenney and buy a pair of socks, when they sold half of them, they would send the card off and that card would say, OK, they're getting low on these things. And the computers would tell you what to start building.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 9 of 38

1041 Restoration Oral History Panel

And in one case, there was a little over $1 million order got replicated because they thought they did a retry. And so they wound up with two $1 million orders. Of course nobody knows that. It's just that the trucks backed up, all this stuff gets built, and they start delivering it to JCPenney, and they accept it all.

And then the computers start getting into an argument over how much you owe me. And so Burlington says, you owe me $2 million plus. And JCPenney says, no, we only owe you $1 million on this particular bill. And so I worked on that. They didn't want that to happen again. So they wanted to put in safeguards. Kind of like the FBI safeguards they put in to keep from duplicate errors and stuff like that.

And that was a fascinating problem because you weren't there just to fix the trouble. Because it could reoccur at any time. You had to fix the software, then you had to fix the customer's problem so that even if it did occur, they would catch it. So those are just some of the problems that we had to work on that were a long time. The problem failed once a month ago and now you're told you've got to go fix it and make sure it's fixed.

Garner: So what year did you retire from IBM?

King: I retired from IBM in 1991. And like George, they offer you a window and you talk to your kids and they say, do it. And we've been kind of broke ever since, but it's been a fun ride. And then this thing came along and start working with the same personalities that you worked with all your life.

Garner: Get your hands greasy again.

King: Yes, get your hands greasy and you get guys who can fix almost anything.

Garner: OK. All right, thanks, Frank. Allen, do you want to go next? Say who you are.

Palmer: My name's Allen Palmer and I quit high school at 16 in New York City. At 17 I joined the Navy. And fortunately through some advice of an older cousin, I paid very good attention on the day they give you a test. I had always in school studied math, science, and taken all my majors. And so they sent me to electronics school.

And when I was getting ready to get out of the Navy, I joined '55 and in 1959 I was getting ready to get out. I didn't know exactly what I was going to do. And my mother who worked for a dentist in New York City always admired two groups of their patients. One were FBI men. They came in with their suits, very formal. The other were the IBM men. And she said, try IBM.

So I was in San Diego and I went down to the IBM office. And at that time in April of '59 there was a little mini recession going on. And IBM was not hiring. So I went back to New York. And sometime, I'm going to guess about August or so, September, there was an ad in the paper to go down to Manhattan to IBM headquarters. They were looking to hire.

And I went in that Saturday and there was a big room just filled with guys. Obviously there were guys at that time. And they would call a group in and you'd take some tests and then they would call out the names and they would thank those names and tell them to go home. Until the end of the day there was a small group of us.

And all I got was an application form. I'm not exactly sure why. I thought maybe, well, they're not hiring me or whatever it was. So I didn't do anything about it.

And in November of the same year, I was working for a company, a small company going out on the railroads and putting on hotbox detectors. So I saw this ad for IBM again out in Long Island where I was living. And I went in and I applied and the guy said to me, have you ever applied to IBM? So I explained what had happened earlier in that summer. And he said to me, do you realize that they are probably still waiting for you to send that? That is the hiring [offer letter]. I didn't understand that.

And he said, well, I'm with SAGE. And we are looking to fill a new class. And we just finished a class in Las Vegas. And this class is for Minot, North Dakota. I had my wife who was pregnant and one child and we were not going to Minot. And he said, the guy in the next cubicle is hiring for Poughkeepsie. So I went in there, got hired, started out in the end of November in Poughkeepsie. And when I got to Poughkeepsie, you have to remember, this was November of 1959.

And when I think about my career at IBM, I look at many aspects of it. And one of it, when I look back just in awe, is the technology changes. Because when I started November '59, they had hired a whole group

CHM Ref: X7542.2016               2015 Computer History Museum               Page 10 of 38

1041 Restoration Oral History Panel

of ex-military electronics people because they needed a massive number of people because was what was happening in the industry at that time, I mean, within IBM, the growth. And we went down, and at that time we were learning how to read IBM schematics.

Also we were testing tube racks. It was a tube [pluggable unit], if you've never seen it, it's a metal frame. I think it has eight tubes on it and a little point and a handle. And there's a tester. We sat in front of that tester. A few weeks later they replaced them with an SMS tester. The SMS card had come out.

And after about two months learning IBM's diagrams, we were sent to different departments in the Poughkeepsie plant. And I would set up the tape drives. And the tape drive manufacturing process was a two part process. Down on the first floor is where they were physically put together, just physically put together.

Then they were sent up to us. And we did all the electronics. We got rid of all of the bugs, the wiring bugs on it. We did all of the timings, all of the settings, everything else. And then they would go out to the floor. And Poughkeepsie at that time was building 7070s, 7090s.

I can remember there was stretch, IBM 7090 StretchS, which was the American Airlines reservation system. And what would happen is at that time they did what I would call a “factory marriage.” All the pieces, tape drives, punches, everything, was there on the test floor out there with the company's name hanging from a big plaque. And they tested everything.

Garner: Must have been a massive test floor.

Palmer: It was huge. And a unique thing about it, there was a crew that came in at night in the tape department and other departments. They were so busy they were working two shifts. There was a crew that came in at night. And what their job was to do was to replace pieces of the hardwood flooring.

The whole manufacturer-- that second story. I can't remember the first. Was all beautiful hardwood floor that IBM maintained meticulously to every single piece. It was huge, huge. And in '61 I wanted to go back to Milwaukee with my--

Garner: Just one second on that. These are 729 tape drives, the same ones we have on the 1401?

Palmer: At first we started with 727s. Then we went to 729s,s which were the first really transistorized ones, or discrete SMS ones. Then I went out to Milwaukee in '61. And I went to work in the customer engineering group, CEs. And that's what this CE name was, Customer Engineering.CE

And again, thinking about the transition of IBM through over the years, at that time the FE's were not a division, the field engineering group. They actually, the branch office was run by the head of the sales department, the manager of the DP office. And even the service people eventually reported to him.

Eventually what happened is as IBM grew, and it was growing so fast, the hiring and service division became more and more important. It was made its own division and O. M. Scott. Remember O. M. Scott? O. M. Scott was the first president.

And so in '61 when I went out there in the summer, I went up to Rochester, Minnesota for basic school. That was hard work. Key punchers, verifiers, sorters, printers, 402s, 403s, 557s. And about two or three years later, what happened was when the 7000s were coming out, the sales department was composed of salesman and systems engineers. SEs. And the SEs basically were going out. And you could have a team of SEs and an accountant that worked there full time.

Garner: I mean, 7000 was a very expensive computing system. So they probably have to staff 20 or 30 people for each installation.

Palmer: Well, depending upon the size of the installation and the activity going on would be how many SEs would be assigned to it. And they were there all the time. But what the change was that from these big accounts, the 7070, 7090, 7094 accounts, to when 1401s hit, you came to small accounts.

I had a [1401] account that was a small shoe manufacturer. Another one was a paint company, made paint. This huge impact, the sheer number of accounts that were coming out there. There weren't SEs to go out and spend time with them or to fix all their problems. So what they decided to do was to take a certain number of CEs, hardware CEs, and send them to programming school.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 11 of 38

1041 Restoration Oral History Panel

So first I went to 1401 school up in Rochester. Then maybe a year later I went back to 1401 school. I mean the 1440 school. They taught it this way. Remember how the 1401 worked? 1440s the same except for. That was it. It was really a fast school.

Garner: What year was that?

Palmer: I'm going to say that was probably about '65.

Garner: The 360 was about to come out, so that must have been the next thing they threw at you.

Palmer: Well no. Because what happened is I never really went to a 360 school. What happened was, this programming that we would do, this program support that we were doing, and that term becomes the name of the job we did, was part of our hardware jobs.

Because from IBM's point of view, what we now had was a person who knew the hardware and the software and can figure out which was failing and you took care of your accounts with hardware and software guys. Well, it grew so large that they went and made a career path. They called it PSR, Program Support Rep. And they had to go for a while. And then they actually went made that position nonexempt. Now a few years later the government came back and said you can't, and they had it go back.

Garner: Because you guys were working all crazy hours again, right?

Palmer: Well, both hardware and software you were always on call. And in my branch, the Milwaukee branch, you were assigned the weekend. You were expected that weekend--

Garner: Any time.

Palmer: Any time. And there was no pay for being on standby. Now, if you got called out, IBM was nice. They guaranteed you at time and a half two hours and 40 minutes, which was four hours of pay.

Garner: Do you remember any harrowing experience for the customer support?

Palmer: No.

Garner: Managers breathing down your neck, get it fixed in 10 seconds?

Palmer: Oh, yeah. We used to always say the salesman's job was to take the customer off and buy him coffee, keep him happy, and we would fix it. The things I remember the most about my career in IBM were more the human side rather than the hardware side or the software side. I mean, you had a lot of instances. You always had problems you think back to fix.

Garner: Before you start that, what fraction of the problems do you think were hardware versus software? 50-50?

Palmer: The 1401 was an incredibly reliable machine. A territory was built. IBM built a territory on a certain number of hours. And every machine was given a number of [CE] hours (CEHs). A key punch maybe--

King: One hour and two tenths [per month].

Garner: That's how often you have to go by on average?

Palmer: On average, that's how much time you should spent on that machine. And IR [incident] reporting if you had more than that, they'd want to know why. And so actually for the physical size of machine, the complexity of the machinery, the 1401 was an extremely, extremely stable machine.

The 1440 had one problem, and that was its electric key punch and this electric typewriter over there. And the tilt and rotate tapes. I mean, you had this electric down, the system was down. And I had a friend who worked for the OP division, Office Products division, and I got him out more than once to fix a typewriter.

But the things that stick in my mind, if you're interested, are more the human things. IBM had a program called Speak Up. And what the Speak Up program was a form, it was in every office, and you wrote down whatever your problem was. And you mailed it into a special, secret office and they took it and they tore it in half. I'm telling you the truth. And they separated it. And they kept all of the personal information as to who it pertained to and they sent the other half off to some executive who would investigate it. And they would come back.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 12 of 38

1041 Restoration Oral History Panel

And I had a medical problem with my eyes that would cause me to lose anywhere from six to eight, maybe nine weeks, twice a year. And one time my manager called me in and he said, you notice you've not had a pay raise in a long time and no promotions. And that's because we have to have other people help carry your territory.

And I did a Speak Up and actually called them up. And the next day he came up, we met off site, explained it to him. And he went back and interviewed my manager and branch manager, and went to IBM headquarters. And the answer came back that an employee will be appraised and rated on the work he does on when available to come to work. It was a very, very big decision that they made.

Well, what I did was, as a PSR, as a software PSR, Program Support Rep, I got very involved in CICS, Customer Information Communication. It was a data processing communication system. And in 1977, I got called into my manager's office and said, listen, they need some help out at level two in San Jose on CICS. So I came out for a couple of weeks. They offered me a job and I stayed. And then I was in the CICS group for a while and then became a manager. And I managed in that group. And I was over in the IMS group. No, I was managing there.

And one of the departments that I was managing was DB2, before it was announced. And in IBM, the software support structure that time was unique in the fact that the developers down on the Bailey Avenue site there, when development developed, announced, and shipped a software product, they no longer had any responsibility for the maintenance. That fell into the FE software.

Garner: They threw it over the wall.

Palmer: They threw it over the wall.

Garner: Take it. You take it. Good luck.

Palmer: [INAUDIBLE]. It went out with bugs. And we used to have to make correction called PTFs, temporary program support fixes. Anyway, there was a product, DB2. And it was going to be announced the very next day.

And there was a big meeting, took place in San Jose with service planning and everybody else. And a service planning manager, second line said to me, “Don't you say a word!”!” And on this side of the table was a vice president FE. And his whole staff going on down.

And a presentation was given about “Oh how wonderful this was” where I was sitting in the back knowing that it had a severe design problem. You couldn't fix it. And all of a sudden-- you won't believe this-- I jumped up, went to the front of the room, and said, “You want to know the truth? I'll tell you the truth. It doesn't run. It cannot ship!”!”

Tom [Vasalaties?] who was the vice president, was sitting at the table, looked over to the second line manage [INAUDIBLE] and said, “Is what he said true?” And I've never seen this before. He had a presentation book in front of him. He took a group of papers, rolled it up, stood up, leaned over the table, hit him in the face with it.


It [DB2] didn't ship for a year.

Garner: So you’re perceived positively then after that? You weren't sliced and diced and shown the door?

Palmer: It was evening. I ran back up there where my office was to a friend of mine who was the other manager. We ran off to the local bar. We sat there and we're having a drink. The third line from San Jose came in and said, OK. Then I was given an assignment to lay out what it would take to get the education.

Garner: To fix it.

Palmer: Well, fixing it had to go back to development. And a couple of weeks later I had to make a presentation on how we were going to handle this problem. And Tom [INAUDIBLE] found me, I was down in Tucson. He stuck his head into office where I was preparing my presentation and he said, you still have a job, don't you? And so to me, I found that--

Garner: That's the Speak Up thing, right, you spoke up?

CHM Ref: X7542.2016               2015 Computer History Museum               Page 13 of 38

1041 Restoration Oral History Panel

Palmer: Well that one was just a shout out, stand up and everything else like that. But after that, I was managing the IMS [Information Management Systems] department. As the manager. I had to make a presentation to a second line manager coming in. And he asked a very powerful question. And it was an extremely good question, would have told him a lot about it. And I sensed he didn't quite understand it. His name was Darryl McReynolds.

And I gave an honest answer, but not a full answer. That evening my boss came to me and said, Darryl wants to talk to you about a job. He came in my office and he offered me a job to go to work for his staff at FE headquarters. And I said no, I wouldn't go to work. Because one thing, I didn't want to move. And two, he didn't know the business.

And I went to the blackboard and diagrammed what he should've expected for an answer. And he said, that's why I want to hire. And then I ended up going to work for his boss, George [INAUDIBLE]. And being in charge for money and how many people staffing.

And in 1991 they came out with a buyout package. And it was very easy for me to leave IBM. Because at that time, I was fed up with it. Because what had happened, I would say from about 1988 to 1991 IBM was in very desperate condition. John Akers was the last IBM CEO inbred at that time.

IBM had all these programs coming down. Six sigma, we were going to get better. 10 times better every year. And you go back to headquarters and you try to explain to people that it's still an odd software. It's still written. There's a lot of problems out there that were out there from years ago.

You're not going to be able to kill service. They want all these programs to just make people work harder and harder, faster, faster, work with less, do more, get smarter, it's your fault. So when 1991 came along and they made the buyout package I had 32 years, took a year's salary, signed my papers, walked out on a Friday night and never, ever went back.

Garner: Well done. OK, thanks, Allen. All righty, Bill, do you want to go next? Say your name.

Bill Flora: My name is Bill Flora. I was in the Navy for four years.

Garner: Speak up just a little.

Flora: And in 1959, I got out of the Navy and went into the SAGE program. Several people have talked about SAGE. So I started in the Kingston, that's where the classes were for Sage in my case. And in October SAGE training was complete.

Garner: So just a little context. SAGE was a massive computer. Size of three building floors. And how many CEs would try to keep-- and all vacuum tubes. Tens of thousands of vacuum tubes. So somehow you guys were supposed to keep a monster like that working?

Flora: That's true. Actually I was in the I/O section. The training for the CEs was I/O people, DISPLAY people, and the central computer system. So after the training, which ended in October, I believe, I went to Corvallis, Oregon for the Portland Air Defense sector. So what this SAGE site did was it had radar input into the computer looking for hostile targets coming to the United States. And we had the SAGE sites all layed around in the [US] perimeter.

Garner: The DEW line was the radar was what it was called. The DEW line, as I recall.

Flora: Some of it was part of the DEW line. But I think that the actual SAGE system was separate. Now, IBM CEs were maintaining this. But the Air Force was doing the operations. They were the people that set at the consoles.

Garner: Do you remember, Bill, any challenging problems with the SAGE system itself that you were involved with or was it pretty regular?

Flora: Not really. It was a three shift operation.

Garner: So around the clock. Which shift were you in?

Flora: Well at the time, every month you could switch to a different shift so that everybody had an opportunity to be on the night shift and so on. But then I took some computer classes, or some education at Oregon State.

Garner: About how old were you at this time?

CHM Ref: X7542.2016               2015 Computer History Museum               Page 14 of 38

1041 Restoration Oral History Panel

Flora: 22. So I was only there in Corvallis for a couple of years. And I was given the opportunity to go to Los Angeles. And the idea was to send me to 1401 school. So I went to 1401 school in San Jose. And then I spent a day in Los Angeles where I would be a CE.

And then I went to Poughkeepsie for 7070 training. And then another couple of weeks of 7074 training. And so then I moved to the L.A. area. After being there for just a short period of time, they moved the customer that I was working with from the L.A. downtown office to the Santa Monica office. So I worked at General Telephone Company for about five or six years.

Garner: As someone in their mid 20's, you couldn't complain about being in Santa Monica, I suspect.

Flora: No. It was a desirable place to live. Anyhow, I was servicing the 1401. And we had a 7074 there. And I was servicing that along with all the peripheral equipment.

Garner: Do you remember something of what their configuration was like? Did they have a lot of 1401s? Who was the firm? Who was the company?

Flora: I think we had two 1401 systems. They had 7330 tape drives.

Garner: “Creepy crawlers,” as Allen calls them.

Flora: Pardon?

Garner: The horizontal ones. Allen calls them the creepy crawlers. Not the 729s that we have.

Flora: So they also had paper tape readers. The message unit calls and long distance calls punched paper tapes in the switching rooms. They brought those paper tapes into the account and put all that data on to tapes which the 74 read. And those were 729 tapes. So I think a couple of years later the 360 got announced. And so I went to mod 30 school and the 360 model 65 school. The model 30 system had a 1401 compatibility feature and the model 65 had a 7074 compatibility feature.

Garner: Bill, was your memory of the 360 model 30s, do you think, in your experience, were most of them still running 1401 programs? Could you tell?

Flora: Initially. I don't remember working on a model 30 system that didn't have 1401 compatibility.

Garner: Did you find that the model 30 was more reliable or less reliable than 1401?

Flora: I would say more reliable. Other than the static electricity problem. Eventually IBM put shielding around the C-CROS units to get around this static electricity problem.

Garner: When did you transition into programming?

Flora: Well, in the late '60s, 370 got announced. And I started working on the 370 system. I never got trained on the mainframe. But I got trained on the tape drives and the disk drives and the optical character readers. And that took place in, let's say, 1970. During this time frame, I moved from the Santa Monica branch to the L.A. midtown to L.A. south and then back down to midtown L.A. where I had to take Interstate 10 to work every day.

Garner: It was pretty smoggy back then too, I think.

Flora: Pardon?

Garner: It was fairly smoggy back then I think as well, right?

Flora: I wasn't concerned about that too much. But the commute was pretty bad. But then I had the opportunity to go to the aerospace building, which is close to the L.A. airport. And there I-- let me see what I was doing there initially.

Garner: Did you have a family by then?

Flora: Oh yes.

Garner: Now did you have to be on call 24/7 as well?

Flora: Yes.

Garner: Anytime of the night you might get a call?

Flora: Yes.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 15 of 38

1041 Restoration Oral History Panel

Garner: So the family must really appreciate it then.

Flora: I don't recall that we were scheduled to have coverage for a weekend or anything like that. But I got promoted to advisory specialist. And that was an exempt position. When I got called out in the middle of the night, I didn't get paid any overtime. But I was able to take time off to compensate for the work. In the '70s is when I became a PSR. And I was working with banking customers in the L.A. Area. And eventually I went back to the aerospace building as an instructor. I was teaching basic OS training to other PSRs.

Garner: The OS 360?

Flora: Pardon?

Garner: OS 360 or OS 370?

Flora: OS 360 initially. Anyhow, then MVS got announced. IBM so I went to Chicago. The Poughkeepsie had put a class together for what was going to be the CE education PSR education. So I went to Chicago and sat through this MVS class.

And then was given the opportunity after Poughkeepsie had given us this training to take a section of the training and teach that for the second class that was done. And then eventually we took that education package back to L.A. and taught MVS PSRs for I think about 2 and a half, maybe three years.

Garner: Now Bill, when did you retire from IBM?

Flora: I retired in 1993.

Garner: Now, I think you were working as a consultant after that, right? Because as I recall, when Y2K came along, you were looking for old COBOL programs to see how they might have a Y2K problem.

Flora: No.

Garner: No?

Flora: After I left IBM, I took a year and a half off and then I got a job with a health care software company. And it was during that time frame that we were concerned about the Y2K issue for hospital.

Garner: Did you find any examples in the COBAL of a real Y2K problem? Or was all the coast was clear?

Flora: No.

Garner: Did not find any problems?

Flora: When I was let's say four years prior to my retiring from IBM. I'm sorry.

Garner: You're thinking of a Y2K issue at IBM? Or COBOL?

Flora: I worked in the language product area. So I worked on BASIC, APL, FORTEANFORTEAN, COBOL, but not that much in COBOL. So in 1993, was given a window. And I took it.

Garner: OK, great. All right. OK, thanks, Bill. Iggy, you want to go next? Say your name and about when you joined. You guys all joined about the same time, all these young 22-year-olds except for Allen.

Palmer: I was 21.

Garner: You were 21. Yeah, same thing.

Ignacio Menendez: My name's Ignacio Menendez. They call me Iggy. My electronics started with the Air Force. I went to the Air Force in 1961 from 1965. And I got trained in basic ground radio radar and then I went into telemetry and spend basically all my career at Vandenberg Air Force base working with telemetry station for the ICBM Minuteman missile. We had all those test firings going down range. And the way we would take all the data coming back from the flight. And it was all processed by a custom computer.

Garner: Custom, not IBM computer.

Menendez: No, no, no. It was computer made by Radiation Incorporated and Autonetics, part of North American Aviation. And we just basically decoded all the flight data. It was all multiplexed and one single

CHM Ref: X7542.2016               2015 Computer History Museum               Page 16 of 38

1041 Restoration Oral History Panel

data stream. And we would receive it, un-multiplex it, decode it. I made a lot of reporters and script charts and whatever out of that.

Garner: So you would actually do reports on altitude and speed and all that?

Menendez: All the, yeah. We had 756 channels of data coming from the flight.

Garner: I have to ask the question, did one ever blow up? We just saw Gravity [movie].

Menendez: Missiles would blow up all the time in the beginning.

Garner: I didn't think that would be the answer.

Menendez: In the beginning, yeah.

Garner: Well that makes for less data to collect in that case, right?

Menendez: Well yeah, but at least from since you're transmitting the data, you don't have a black box like an airplane. You have live data coming in to a point that the missile blows. And in fact, we also had the responsibility of the command destruct system, which is what blows up the missile.

Garner: Did one ever head to L.A. You probably can't tell me. That would make a good story.

Menendez: No, they would never get that far. What happens is there's a range safety officer that has this visual thing kind of like a little window. And as the missile takes off, if it deviates at all from that side, command destruct is activated, and the missile's blown in the air.

And it never really gets that far from Vandenberg when they blow up. So anyway, that was my background. And I worked in that of my entire Air Force term. And when I was ready to come out, I was interviewed by IBM.

Garner: What year was this?

Menendez: 1965. And IBM came to the base, gave me an interview, then I didn't hear from them for a while and I thought, well. About two months before I got out, they called me and they said, when will you get out? I said I get out December 10, which was a Friday. And they said, would you have a problem starting December 13, the next Monday?

So I went to San Jose and started working in San Jose on a machine that's kind of infamous called a “noodle picker.” IBM 2321. I worked on final test initially. Then I worked in system tests. Then I worked on a QA function called a new product line acceptance, NPLA.

Garner: What were your feelings about the noodle picker? Everyone has an emotional feeling about it, because it was a very complex mechanism, hydraulics, air compressor, ...

Menendez: I tell you what, it was really the only way they could get a lot of data. Because the 2311 disk could only store 7.25 megabytes.

Garner: Yeah, about seven megabytes.

Menendez: [For the] 2311 drive and the data cell was 400 megabytes.

Garner: 400 meg.

Menendez: [Equivalent to] 54 2311s.

Garner: 54 disk drives.

Menendez: In one machine. And the access time was you could get the farthest away you could be was 0.6 seconds from getting the next piece of data.

Garner: Grabbing a strip.

Menendez: [INAUDIBLE] one was ready in the drum, put it in a way, seeking farthest away 180 degrees around.

Garner: Rotating the drum.

Menendez: Rotating. Yeah, rotating the bin that had the strips. Picking up the strip, rolling it past the read/write head. That was 0.6 seconds. That was the worst timing. Now, the best timing was if you

CHM Ref: X7542.2016               2015 Computer History Museum               Page 17 of 38

1041 Restoration Oral History Panel

already had a strip. There was 100 tracks on each one of those strips. So if you wanted just an adjacent track, it was just 50 milliseconds. [INAUDIBLE].

Garner: Dave Bennett also worked on that. Told me he could walk in a room and by the smell he could tell if it was working right. Did you have that experience?

Menendez: Well, it smelled a lot like hydraulic oil. But I tell you what, those machines were very complex. If they were adjusted correctly, they worked really good. And the problem was in IBM was that field engineering never really had a good training for the CEs for the 231.

They would try to train them in five days. There was no way they could really master that machine in five days. When I was in the field, I ended up servicing-- I transferred from the plant to field engineering in Palo Alto. And I ended up servicing all the machines in all of the Bay Area.

Garner: All the noodle pickers?

Menendez: All the noodle pickers. Right. Amongst the other 360 I/O. And what happens is that it was just a matter that the CEs just didn't understand how to adjust the machine right. So I ended up going and overhauling the machines. Just everything happened and they would run again for six months again and then it would go fall right back into the same thing. Because the machine would wear and you had to really adjust them all the time.

But I had an account in Palo Alto which I guess remains nameless. But I had 24 noodle pickers running at the same time amongst two model 65s. And I wouldn't get a failure more than maybe once a week amongst all 24 drives or an individual 2321 once in six months. So they really run good. When they are adjusted right, they run very good.

Garner: Was that about same reliability as a 2311?

Menendez: No, no, no. Not at all. The 2311 was just strictly an actuator moving.

Garner: But they would fail less often.

Menendez: Yeah. Those would fail less often. But I had them running pretty good, actually. And I had other responsibilities in the branch besides the noodle picker. I was also the specialist for a branch for the mod 50s, 155s, 158s.

Garner: These are all 370s.

Menendez: Well, 360 model 50, then 370 155 and then the 155 model two with that unit.

Garner: For the I/O and the CPU?

Menendez: Oh yeah. Yeah, yeah. 3330, 3420, 3211, 3705.

Garner: Now the 3330, is that where we got started with thermal conduction modules?

Menendez: No, no, no. The TCMs came on the CPUs on the 3081, much later.

Garner: Did you ever deal with those?

Menendez: I dealt from the San Jose site part of it. What happens is I was working on service planning hardware, which was also called TechOps back then. And they had a design problem with 3081. Well, a design characteristic that if it really get heavy loaded, they have two processors going for the same cache. And there was no tiebreaker and they would just deadlock the machine.

Garner: Deadlock the machine. Yeah.

Menendez: And the service processor would come in and figure out that the machine was kind of dead and would start doing a complete checkpoint restart. So what happens is they had all those machines ready to be announced.

Garner: What year was this about now?

Menendez: This was around 1980. Trying to think. '85, '88, something. And what happens, they had this big order from all the consortium of Japanese banks for 3081s. And they said if you don't get the 2305 running on the 3081 and we don't want the 3081s. So what happens is that--

Garner: The 2305 is the noodle picker.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 18 of 38

1041 Restoration Oral History Panel

Menendez: No, 2305, now I've jumped machines on you. 2305 is the drum machine.

Garner: Oh, that was the drum.

Menendez: Which was designed for the model 75 and the model 91, which were the only ones that handled the 2880 channel because of the speed of data transfer.

Garner: It was unusual for IBM to have drums. What was the drum for? It was a communication mechanism?

Menendez: No, no, the drum was for really fast access. Like to put some of the sysres stuff on it, some of your indices. And then later on where they have virtual memory, what happens is they end up using for paging device. And really the system would really speed up without this 2305 machine because they were fixed heads. They had no seek time, just latency.

Garner: So the banks wanted this.

Menendez: Yeah. And what happens is when they were going to go from 370 to 3081 and it was really no gain at all if they wouldn't get the 2305s, because they would have to go to a regular mechanical seek type device to do all their paging and they would lose all the advantage.

So I was in service planning at the time. And I heard about it and I opened my mouth and I said, “Oh yeah, we can fix this.” I saids no, you can't because the machine is old. It was design back in the '60s for 360. And now we're in the '80s. And everybody said it can't be done. I said, yeah, it can be done.

Garner: They challenged you? They thought you would be incapable of helping? Is that it?

Menendez: Well, what happens is I opened my mouth, said that on the meeting, and then I went away. And then about a couple of weeks later, engineers called me back into the meeting in a room full of microcoders that said that it couldn't be done.

I explained to them how it could be done. And they still said it couldn't be done. So about two weeks later, again, my manager calls back and says Microcode Engineering wants me to work for him. So I went to engineering and microcode and I redesigned 2835 control units for a drum.

Garner: Did they embrace you or did they ignore you? I mean, were you able to work with the microcode team? Did they like your ideas?

Menendez: Everybody was against it because I was going to prove that they were wrong. So no, I would get absolutely no help. I had one young engineer that was basically my operator. He would run the 3081 console and do all the testing for me. But I would do all the microcoding. And there was no compiler, no assembler, no nothing. It had to be all entered in the little console of switches.

Garner: Oh, you couldn't even load it through cards or through tape?

Menendez: Not when you were designing something. And once you designed something, then of course you could make a data set and load.

Garner: You had quite a load on your back.

Menendez: But they took me one year to redesign and test and ship that code.

Garner: And it ended up shipping?

Menendez: Oh yeah.

Garner: You were proven?

Menendez: It shipped. The 3081 would absolutely not work without this new microcode.

Garner: Did the other microcode guys ever talk to you again?

Menendez: Well, yeah, they eventually kind of started talking to me. But they weren't very happy that I had--

Garner: Stood them up, yeah.

Menendez: Stood them up. And I had to redesign that code twice because there was a two channel version and a four channel version.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 19 of 38

1041 Restoration Oral History Panel

Garner: So what gave you the inside knowledge to understand? Just because you had worked with it before?

Menendez: Well what happens is I had worked with the 2841s and with model 50s. And I had an account that I had a model 50 every Wednesday for four hours. Stand alone time. And I had goofed around with that horizontal microcode through all my desires and tricks and whatever.

Garner: That was like a scheduled maintenance day, once a week?

Menendez: Yeah, that was with a customer in Menlo Park, Judson research. Part of Allstate Insurance. And that's where they computed all their rates and morality tables and all that. And I had four hours of a standalone system totally for me.

Garner: Your personal computer.

Menendez: So that's why I got the inside of how microcode really worked. And when they said it couldn't be done, well I knew it could be done.

Garner: So where did you go from there?

Menendez: After I finished designing that microcode, then I went for product engineering work for a while in 3880. And then I went to STL for a while. Santa Teresa labs. I went for software assurance for a while. When [they] started cutting down personnel, whatever, well I also worked on the architecture of a 3390 drive. I got the chance to work with that for a while. And then they started cutting personnel.

So I went to STL on assurance. And then they called me back and they said they wanted me back in San Jose to start microcode assurance. So I went in and started working microcode assurance for one of the rack mounted ones. I think it was the 9320.

Garner: A disk drive?

Menendez: Yeah, it was a disk drive on racks with the control units. 39. 9320. Something. And from that I went to the micro assurance on a product that was going to be announced. Can I say the product name? It was called CSTAR.

Garner: I know all about CSTAR.

Menendez: And what happens is this is going to be the one control unit to beat it all. It was going to be able to handle everybody. And I blow the whistle because what happens, they had designed in the real wonderful C++. No goto, no spaghetti code, nothing. And it turned out to be thousands of times slower than the existing control unit.

So I blew the whistle. The machine was never going to make it. And of course I made an enemy out of the entire development team there in building 28. And it got pretty ugly. I finally transferred out to Santa Teresa lab. Many months later they canceled the project. CSTAR just died right there. And so then I stayed there and I worked with some more jobs doing software testing in Santa Teresa labs. And I finally got enough time in IBM that I decided I had enough. 34 years with IBM and I retired in the year 2000.

Garner: All right, Ed, last but not least. Say your name and all that.

Ed Thelen: OK. I'm Ed Thelen. And in this police lineup of IBM people, I'm the ringer. I worked for the competition. But I got to admire IBM people. I had a little experience with them. And certainly IBM equipment, especially the peripherals. So when this group formed, I said “please let me in.” So anyway, these people are organized in life. And I'm the other end of it.

I was born in 1931 and went to college. Didn't know what I wanted to do. Went for four years, got nothing. Quit. Went to work in a machine shop. Got the draft notice. Got in the Army. Got to school for one year and fixing Nike anti- aircraft equipment for two years.

Got out and found that being a techie was-- the techies I had experienced were broke most of the time. They didn't have a car, they were taking trolley cars to work. So I went back to college. And so in 1961, I had a little family going. I got a degree. Bachelor's. Seven years of college to get a Bachelor's of Science.

But 1961 was a wonderful year to get a degree. Defense business was booming. Computers were booming. We were on the way to the moon. They wanted engineers. I had resumes and job offers in all

CHM Ref: X7542.2016               2015 Computer History Museum               Page 20 of 38

1041 Restoration Oral History Panel

the pockets. It was wonderful. … IBM did not interview at our school. So I wasn't even aware of IBM, really.

Went to General Electric. General Electric wanted to get into computer business. It turns out that not all of General Electric wanted to get into computer business. It was a bootleg project. They submitted an order, a bid to do Bank of America make a computer for them. And about the same day that they won the bid, they got this letter from corporate headquarters, do not make this bid and don't get in the computer business.

Well, these wild hairs got the bid, made 32 machines for Bank of America, called it ERMA. And they were off and running without the deep pockets of New York corporate General Electric. So anyway, I got hired in. (And my memory isn't too great sometimes and I need cheat sheets.)

So went to work on a little binary machine called a GE-225. It had a 20-bit word length. Three characters per word. And actually was sort of designed to compete with the lower-end machinery like a 1401. And it's my theory that anybody can make a processor. You get some tinker toy circuit sets, make a processor, core memory, da, da, da, anybody could do that and make a reasonable machine. But these IBM guys had good peripherals. Good, solid, peripherals.

Garner: You mean IBM did.

Thelen: IBM peripherals. Good, great card reader. Great printer. Great mag tapes. You name it, it was good. Except possibly noodle picker took a little maintenance. So General Electric being on the cheap side, because they were a bootleg project, had a horrible card reader from Great Britain from Elliott Company.

It resembled one of these meat slicer things, really. And talk about need adjustment. Not every six months. Seemingly every two weeks you had to take the thing apart, get the greases all distributed so the mechanical damping was correct. But I did get a very good schooling there at GE .

And it was really interesting. When you're in sync with the instructor, wonderful things happen. Every word he said rang in my head. I nailed that machine. So we went out and wound up at a site. And we had a floating point [unit]. You guys don't have floating point. We get an extra box the size of two refrigerators to do floating point.

And so we were doing scientific work at a company called Air Products. We were on the second floor doing scientific work with the floating point. The 1401 was on the first floor doing business. And I was getting teased a lot. Here I was full time, fighting these peripherals, and I'm going to get into that a little bit, and people downstairs would say, “Is this is your full time here, what are you doing? Sleeping all day?” I would reply “No, I'm busy.”

And they'd say, “Well, our 1401 guy comes in every Friday, dusts off the machine, and leaves.” And I was up to my ears in problems. And I'd like to discuss some of those problems, because these IBM guys really don't know what crappy peripherals are.

Palmer: No we don't.


Thelen: So we had a printer that was made by Analex. The machine was an Analex printer with a General Electric controller. Now, things are supposed to be operator friendly. Well, General Electric didn't know anything about computers, at least at the manufacturing end. When an IBM printer runs out of printer paper, it's the last line on the form that says, “I'm out of paper,” do something. And you can put in the next form and go.

Well, [when] General Electric and Analex printers found you’re out of paper, the sensor was way below the bottom of the page. So you're about the middle of the page. They detected you're out of paper. And that is not good, especially if that “Out of Paper” does the reset to the controller and you lose the next line of what you're going to print.

So if you have a major print job and run out of paper, the only way to do it was to take a box of paper and another box paper and another box paper and tape the fronts and the ends of these four boxes of paper together and then try to do your run. That's printers. Oh, oh, oh, I"m sorry. I hope I don't wander too much. So these [1401] guys had beautiful print, had a chain printer went this way.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 21 of 38

1041 Restoration Oral History Panel

Garner: IBM.

Thelen: IBM. You guys are IBM. OK. So we'd have this problem that in our printers, dispersion was vertical. And nobody wants to see wiggly lines vertically.

Garner: Get sea sick trying to read it.

Thelen: Yes. On the paper, this E is too high relative to that T and the Z is low over on this other column. Now, these [1403] people have really the same problem in timing the hammers, but it's horizontal. So you get nice looking print this way and nobody's ever figured out how close to put with the D with the Z. That's a drafting problem. And it is not apparent to the eyeball. That's printers. Frank's [1403] printer. Oh golly.

We did have a card punch. It was an IBM card punch. A 514. They leased it from IBM and didn't have a service contract. IBM equipment was so reliable that the card punchers on the GE-225 didn't even have a service contract with IBM. And it worked. Then we can get off into mag tapes. Magnetic tapes, these things that spin around and around with tape going in between.

Garner: You guys used Ampex tapes or something?

Thelen: Yeah, we used Ampex. And Ampex tape drives, those were not designed for business, but for instrumentation. Maybe for missile work. And they used-- yeah, I'm sure you're familiar with Ampex tape drives. And so you'll go click, click, click, click as you're recording this stuff. And life is pretty sedate.

But in business data processing, you're doing rewinds. If you write a bad record, you have to go back and rewrite it because there's error sensing. And so the tapes drives are really taking a beating. If you do a sort for JCPenney, they'd have seven tapes of input data to get started. And you'd be sorting all night.

Garner: Because there'd be problems.

Thelen: Yes. On these drives that were designed to go tick, tick, tick, tick, tick, tick in one direction. So anyway, I fought that for two years in the field, two different sites. I had a secret advantage.

The customers I had wanted to use the machine in the daytime. I would sleep. And at night I would come in and make sure the equipment was still working right and program all night long. It was my machine. And oh, just play time. And if the machine broke in the daytime, you're off like a fireman.

Garner: Is that where you wrote your first pi program? At night?

Thelen: Say again?

Garner: Is that where you wrote your first program to calculate pi?

Thelen: Yes, yes. I'm a pi computer fellow. OK. Pi is an interesting thing. You have to worry about overflow if you want to compute thousands of digits. So you have overflow problems, under flow. And you really learn the machine, the internals of a machine. And that was what I wanted to do with that.

So anyway, I got to improving the diagnostics that General Electric had. I made a better tape diagnostic that would exercise the tapes harder and show faults. And eventually I got a job at headquarters. And I really wanted to go down there and tell those headquarters drones that field people had problems. But the drones didn't want to hear that.

Garner: Are you being polite calling them drones?


Thelen: I claim that if you've never been in the field and go to a headquarters, you get headquarters-itis, which is a really bad disease to have, but you don't know you have it. But field people know you have it. There's a reality to try to keep the customer happy that headquarters people really don't get. So after a while I realized that General Electric-- oh, I got into time sharing a little bit. I helped some friends who were converting GE/Dartmouth Time Sharing into GE Commercial Time Sharing.

Garner: Time sharing on the GE computer was extremely slow, as I recall.

Thelen: Well, we used two machines, a quick DataNet 30 to service the TTY lines and a main computer to compile and run the programs. It started out a GE 225 and it had a 21 microsecond memory access time. That's pretty slow.

Garner: 21?

CHM Ref: X7542.2016               2015 Computer History Museum               Page 22 of 38

1041 Restoration Oral History Panel

21 microseconds!?!?

Thelen: 21 microseconds memory access time.

Garner: That's really slow.

Thelen: And so yeah, it was fairly slow. Soon they replaced the GE-225 with a 235 that was three times faster. And then they migrated the time sharing to the 600 series, which is one-to-two microsecond memory access time.

So anyway, I decided that General Electric would not sustain their losses. Why were they losing money? They were not getting customer referrals. People bought IBM equipment. You paid more, but you were assured of good service and equipment. For a long time I never heard that any GE customer ever reordered General Electric computer equipment.

Then I heard that the Army testing boosters and missiles had reordered. But they didn't have much peripherals except for paper tape. It was a data logging operation, with punched paper tape output. And eventually I heard that there was a bank in Vancouver that reordered General Electric equipment. But you can't make a living that way. General Electric was losing money. I bailed out.

I heard that IBM was hiring. And so I joined IBM, their time sharing system effort. Bright eyed, bushy tailed, I figured I could help. Well, two things. One, they were in deep doo doo. They had a basic mod 65, 360/65. Two microsecond memory access time. And they wanted to do a virtual machine as well as time sharing. And the hardware just couldn't keep up with the dreams of the software people. And I hauled the family all over the country.

My middle kid said I was an itinerant programmer. And the wife really wanted to go back to Minnesota. And those were the days. You had job offers out of all your pockets. So I called up Control Data that had given me a job offer and said, “Hey, is that job offer still open?” “Yep. Where and when do you want the truck?” You know? Those were glory days. So I want to work for Control Data Special Systems. Worked on their 6000 system that was a world beater.

Garner: 6600 you mean?

Thelen: Pardon?

Garner: 6600, the CDC 6600.

Thelen: 6600. Did I say wrong?

Garner: No you said 6000.

Thelen: I beg your pardon. 6600.

Garner: That was the computer that was embarrassing IBM, actually.

Thelen: Yes that was the--

Garner: World's fastest scientific machine at the time.

Thelen: Yeah, that wound up causing a lawsuit. IBM was pretty competitive, offering vaporware.

Garner: Watson said, how can 34 people, including a janitor, outperform IBM?

Thelen: If you have a genius called Seymour Cray. That's the way to do it.

Garner: Did you ever meet Seymour?

Thelen: Never did. I tried twice. Never got to see him. So we had a good time. One of the customers for special systems was a telemetry thing for flight testing the Grumman F-14 Tom Cat. And I know all about IRIG inter-range. I forget what IRIG stands for now.

Menendez: I don't remember now.

Thelen: Inter-range something. [Inter-Range Instrumentation Group]. But Seymour Cray had left Control Data and they got lost. Control Data got lost. It was sad. So I wanted to go to a place that knew how to make money, especially in software. We were making good software for one customer.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 23 of 38

1041 Restoration Oral History Panel

So went off into process control to work for a company called Measurex. And was there for like 18 years. Had a good time. And eventually retired. The young kids are just much faster than the old duffers. I hate to say it. And I could retire and I did. I didn't want to have to compete with these young whipper snappers.

Garner: What year was that?

Thelen: That was about 1996.

Garner: All right, thank you.

Thelen: So I got it in the computing bracket in 1961 and left in '96.

Garner: All righty. All righty, thank you, everybody.

Palmer: Can I add one thing?

Garner: Sure, yeah, absolutely. Allen always has the last word.

Palmer: I thought someone else would bring this topic. And if you don't mind, I'd invite anybody to add to it over here. I think one of the things that was not mentioned today, particularly in asking our careers in the field, was an understanding of the philosophy that IBM had about ownership of your territory.

Out in the field, you weren't just assigned any random customers as the call came in. You had a set of customers. And from the day you were hired, you were always taught and just instilled that you were responsible for them as if you owned them. And you were measured by their customer satisfaction. And so people took great pride in running their territory, knowing what to do, making sure that everything was taken care of for customers over there. It really was a mentality that I think it--

Garner: Hasn't been repeated since then, perhaps.

Palmer: No, but I also think it was a little difficult for the museum to adjust to, to appreciate it when we came here.

Garner: Well many of you guys had top secret clearances. You were brought right into the bowels and heart of the company, right? Their most secret. Companies were running on IT, right? You had to come in and keep them going.

Palmer: When a group of CEs worked for their field manager, they would help one another out. They would get together maybe at lunch time to discuss the problems. You could always depend upon someone to come in and help you. But this idea that you are responsible.

So if a customer complained, your manager didn't look at it as if he was complaining about IBM. It was: “What was the problem? Why weren't you handling it? Were you doing your PM? Were you going out and talking to the customer?”

You're supposed to talk to them. You're supposed to talk to the data processing manager. You were supposed to know your account. And a smart salesman always went in close tabs with the CE. Because the CE would know if [what] they were talking about maybe needing another piece of equipment, if there were changing roles, if there were problems.

And so the feeling of ownership, at least I know that myself and I'm sure all of us over there that were ever in the field, had that idea and that feeling that we owned our territory. Very important philosophy that IBM taught you.

Garner: All righty, this is part two of our 1401 restoration team interview today on Wednesday, July 1, 2015. And I'm Robert Garner. And I thought what I would do is start with just how I got involved briefly as introduction, and then we're going to have an open discussion about the restoration project itself. Over the past, what, 12 years now.

So in 2003, [the] Museum heard of a 1401 being offered on Germany. And no one bid for it. But it was decided that if the museum were to purchase it, the question was asked, who would volunteer to lead a restoration project? And I had just joined IBM. And I didn't know anything about IBM computers. I had worked for companies trying to kill IBM my whole career, which is a big challenge. I'd worked for Xerox and Sun Microsystems.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 24 of 38

1041 Restoration Oral History Panel

And I raised my hand and said sure, I'll volunteer. And then I said, OK, what's an IBM 1401? I knew absolutely nothing about it. I had no idea how big it was or how small it was or anything. So the next challenge was how to find somebody who did know what an IBM 1401 is and how to restore it.

So someone suggested to me, they said there's this thing called the IBM San Jose Retirement Club and newsletter. So I crafted an advertisement to put in that said:: “IBM 1401 Needs Help.” It was published in a month and then two weeks later about a dozen to 15 people showed up, including all you guys. Well, not all you guys, but most of you guys.

And we started the project. The machine arrived in 2004. The German machine, which you guys will talk about, was a bit more of a challenge than we thought. It took three or four years to come up. The Connecticut machine came in 2007. I got a call when I was driving through on the highway. They asked: “Are you interested in a 1401?” And we had challenges with the German machine, so I said “yes!” and purchased the Connecticut machine at that time.

And then we had the 1401 reunion in 2009. We remodeled the room in 2013. And it was open to the public for public exhibits in the fall of 2013. So that's kind of the background for the restoration project. Now I'd like to open the mic to you guys. When the first team that showed up, I have everything the names of all the people, some aren't with us.

There was Glenn Lea, there was Dan McInnis, there was Donald Cull, there was Dave Cortesi, there was Milt Thomas, there was Bob Brubaker, Glenn Furlong, there was Allen Palmer, there was Ron Williams, there was Jack Parker, there was Don Luke, Harry Davidson, there was Frank King. Did I mention Chuck Kantmann's name? There was Chuck, there was Bill Flora, there was Bob Erickson.

And we basically organized such that Ron Williams was in charge of the CPU. You guys each had an area of specialty, which I recognized early on. Bill, Bill Flora, we put you in charge of the 1402. Frank, we put you in charge of the 1403. Allen, the tape drives. Chuck Kantmann was kind of overall coordinator. Bob Erickson worked on the unit record equipment. And there was Milt Thomas and Don Cull at the same time.

So that was kind of our initial organization. And the museum had never done a restoration before. This museum has focused most of its energies on preserving equipment, getting it in an accession and described and stored. And here we were trying to get a machine to bring it back to life. They had some experience on the IBM--

Thelen: [INAUDIBLE].

Garner: No, the 1620. And it was a tiny project. And the museum was also starting the PDP-1 restoration. I actually worked on that project for a short time before I came to this project. But on the whole, the museum had-- it wasn't its DNA to restore machines. So you guys, I think, will recall some of the story. So do you guys remember seeing the ad? Who remembers seeing the ad first? Everyone remembers seeing the ad? OK, what was your reaction to it?

King: I remember making a phone call to Chuck Kantmann and he said, yeah, he saw it. But I said, well, they probably are going to go bankrupt anyway. But we'll go listen to what they're talking about.

Garner: The museum will go bankrupt you mean.

King: The museum. They go up and down and through. And I wasn't going to work on anything if it wasn't going to be here forever. So why spend all this money and then they're going to put it in storage and go away.

Garner: And you're used to working for a reputable company your whole career.

King: You used to work in the one that had a plan for survival. Had 100 year plan, not a let's do something in the next four years, which is politics. And he said, well.

So we called a few people and I remember calling Don Cull. And he had seen the ad too. And those were people who were just interested. I just remember we were just coming up here to see what you guys had to say. No intention of actually you passing the test we were going to give you.

Palmer: It was interesting to see it. It's funny because it was relatively small.

Garner: The 1401 you mean, or the team?

CHM Ref: X7542.2016               2015 Computer History Museum               Page 25 of 38

1041 Restoration Oral History Panel

Palmer: No, the ad. It was a small ad. It was not an eight and a half by 11 page. It was a small page down there.

Garner: Two by three [inches].

Palmer: A low-rent ad. And it just seemed interesting. I always liked the 1401 as a machine. And the challenge of it. And it seemed like, hey, let's go off and see what they're talking about. So the ad showed up one day.

Garner: Bill, do you remember seeing the ad?

Flora: I remember seeing the ad and I don't think I-- I don't recall that it had your phone number, I think, and maybe--

Garner: Yeah, it had an email and phone number.

Flora: I thought you had an email address. I think immediately I wrote an email saying that I would like to get involved.

Garner: OK. And then how did you find out about the project?

Thelen: Dag Spicer either emailed or phoned me saying: “Was I interested in a 1401 restoration?”?”

Garner: So he got you involved?

Thelen: Yeah.

Garner: We wouldn't have gotten your presence without Dag?

Thelen: That's probably correct. I would have heard about it eventually, but he got me in early.

Garner: And were you in the first meetings?

Thelen: First?

Garner: First meetings where we had the team?

Thelen: I don't know. I think I heard about it in late May. And there were previous meetings that took place.

Garner: And then you started the website pretty soon. Pretty quickly.

Thelen: Yeah. I do websites, folks.

Garner: And then George, how did you find out about the project? This is the story of--

Ahearn: Me?

Garner: Yeah, neighbors [INAUDIBLE].

Ahearn: I saw that ad.

Garner: Oh, you saw the ad too. Oh, but you ignored it.

Ahearn: And I immediately rejected it. I rejected it because I said to myself, I don't remember very much about that project at all.

Garner: You were one of the [1401] designers!

Ahearn: I don't remember.

Garner: And your memories are excellent.

Ahearn: I mean, the thing is, I felt that if you're going to go and help on this thing, you've got to know something about how to do it. And I had forgotten all that.

King: He's not a field guy.

Ahearn: Yeah. See, that's right. If you lived with this in the field, you've got a lot more experience with it. Whereas if you were in on the early design, that came and went, you went on to the next thing.

Garner: OK, so then how did you find out about it through Bill?

CHM Ref: X7542.2016               2015 Computer History Museum               Page 26 of 38

1041 Restoration Oral History Panel

Ahearn: I didn't find out about it through Bill. What happened was Jud McCarthy in Florida was very interested in the 1401 and so forth. And he told you folks that you've got a guy in your own backyard, namely me, and why don't you get contact him and invite him. This was five years later after you began. You began in 2004. Judd make this contact in 2009. And someone called me.

Garner: Could have been me.

Ahearn: Or emailed me or something.

Garner: But then you learned that Bill was your neighbor.

Ahearn: Oh, I knew Bill was my neighbor, but I didn't know he was involved in this project until the day I showed up and you were all in that conference room having lunch.

Garner: OK. So what do you remember as our biggest challenges from the very beginning?

Palmer: Power.

Garner: Power.

Palmer: Power.


Garner: It was a 50-cycle machine from Germany.

ALLEN PALMER: Yes. And why is it important? Particularly for the tape drives, because the tape drives are synchronous motors which get their speed off of the frequency or the line frequency. And since it was 50 cycles, what meant to say it was that the capstones, which control the movement of the tape, was run at the wrong speed. It wouldn't move tape at the right speed. And the first inverter that you acquired, we were only able to bring up a certain particular piece of I/O at a time. We couldn't ever get the whole system.

Garner: Yeah, it was only the CPU, the 1402 and the 1403 and one tape drive. And that was questionable. And that was questionable. Yeah.

Palmer: What do you remember?

King: Well, I remember that we got enough three phase power.

Garner: For the printer at least.

King: IBM never gave a hoot about balancing the load. And it meant that you could just barely make it if the loads were balanced, which they were never going to be. And I was in favor of converting the machine at first with all those 50-cycle resonant transformers. And I thought we'll never get those working, so let's get a converter.

Garner: So what was your feeling about the 1402, Bill and Frank? It was pretty rusty, as I recall.

Flora: My feelings are it was a big job just getting rid of the rust and corrosion. It really looked bad. But after working on that and that's the way we concentrated in that area is to clean that stuff up.

Garner: But on the card punch though, didn't we-- we swapped with the one in visible storage?? Frank, you remember that??

King: The punch unit.

Garner: The punch unit.

Flora: I don't even remember that.

Garner: OK, well Frank can tell that.

King: We rolled it. We actually took that thing and rolled it in here beside of it.

Garner: We took the one from visible storage into our room.

King: And started. It was the minimum amount of conversion that we were going to have to do. There were engineering differences to the whole thing. As a matter of fact, it's only a year or so ago we actually really got it to punch cards. Or less than a year ago.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 27 of 38

1041 Restoration Oral History Panel

Flora: Part of my situation is I was still working on my second job. And I was only coming in for a couple of hours during Wednesdays. I think it's always been Wednesdays, hasn't it?

King: Yeah, we had a Wednesday morning breakfast in Morgan Hill every Wednesday. So we said, well, let's eat breakfast and go to this museum and it just wound up that all of us were available at the Morgan Hill group. And it was more than just Morgan Hill were available on Wednesdays.

Garner: So what you guys think about the plastic glove thing?

King: The what?

Garner: The plastic gloves.

King:Well, we thought they didn't understand. They hadn't ever done any. And they were right. I'm not saying this as a negative. It was ignorance that we couldn't get together on whether we wore white gloves when you're rebuilding a car engine. I mean for God's sake.

Palmer: Well, it was the antithesis of IBM, first of all. With IBM wouldn't let you wear a ring because they were afraid of getting your hand caught in a piece of moving machinery. She [Allison Akbay] wanted to give us all gloves.

Garner: The feeling at the museum was that artifacts that went into accession You could actually get a fingerprint on them. And if they go on display, there'd be somebody's ugly fingerprints.

So everything the museum does when they accession is they wear white gloves. Now, we were going to be restoring the machine. It had grease and oil everywhere. And they asked us to wear the white gloves and we did it for about-- you guys did it for about a month or two, as I recall.

Palmer: About one month at the most. Well, let's put it this way, we had the gloves assigned to us and then they gave up after about a month.

King: They would come in and monitor us once in a while. And it was funny because they will let a machine come in that's rusty and leave it rusty. Not ever put anything on it to stop the rust from going, because that's the way we found it. So if you can't even touch that with a finger because it might rust. Well, it's going to be a pile of rust in a couple of years.

Garner: For me one of the biggest examples of the challenge at that time was Ron Williams. He's not here with us today. But we were in that early meeting with the museum staff and Ron declared::“You guys have a building full of junk!”!”

Palmer: It was nothing worth taking.

Garner: Nothing worth having, yeah.

King: With security [INAUDIBLE].

Garner: Well, and his reason was nothing operated. It's all inoperable machinery. And so he certainly, like you guys, appreciated the fact that we were going to bring a machine to life. And that would make it valuable. And not so much that it needs to be preserved behind glass, but it needs to be brought to life. Now, did you guys think that we would ever get it working in some amount of time?

King: I never had any doubt about it getting working ever. If we decided. If this group of people decided to make this thing run, it was going to run. But was the museum going to stay in business long enough for us to have done anything worthwhile?

Palmer: I never doubted the museum would stay in business. The thing I remember the most is a meeting we had. And we each took the areas that we would look at. And we each drew up a plan as to how long it would take.

I remember saying, and I still have it written down in the first log book, it would take one year. In one year we'd have tape drives. At this time we had only the German machine. We'd have it all up and running. And it's still down in the log book. I still think about it. And I think, this is the longest year of my life.

King: Well, if we'd had a powerful 60-cycle converter.

Garner: It would have helped.

King: We could have had that computer up shooting bugs on it right to start with. But we never got there.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 28 of 38

1041 Restoration Oral History Panel

Palmer: But that wouldn't have made a difference. I mean, it would have made a difference of getting it up. But as far as getting it done within a year, no. From the tape drive point of view, one of the biggest factors for me was Grant Saviers.

Grant Saviers, who was a member of the board here has a hobby of a machine shop. Which saying machine shop is like saying IBM makes tinker toys. He had the most beautiful machine shop which allowed us to do a lot of rebuilding of tape drive parts. He made tools for punching out [parts].

And I say tools, they're dies for punching out washers, for driving out bearings, getting new powder. And so the original tape drives that came in with the German machine have really, really been totally restored. It was stripped down. And a lot of that is thanks to Grant Saviers.

Garner: Yeah, you guys had reformulate the magnetic powder, the clutches as I recall. At one point you learned that their bearings were angled and you contacted the company and what'd they say?

Palmer: Yes, Grant was big in that. And one of the problems we ran into was that the magnetic powder today is much more efficient. And since you had to fill a physical space, the volume of powder that IBM originally put in there gave you a certain torque curve as you applied powder. If you use that same quantity of powder, you got a torque curve that looked like you had a racetrack. So we had to come up with a compromise. And then a lot of things--

Garner: I recall the bearing though, that company said you could buy the slanted bearings only if we would purchase 1,000!

Palmer: They were $75 each. But we'd have to get 1,000 of them to have them made. And so he found a way to get two separate bearings and a shim in the center to take up the space. It was a lot to ask the museum to buy us about the 16 bearings that we needed.

Garner: So during this early time, we knew that we had to reform the capacitors. And Ed, do you want to talk about your experiences in capacitor reforming? Is it almost a religion, the topic of capacitor reform?

Thelen: Yeah, well, I'd like to go back to power for just a minute. We started out with a converter. People wanted to go rotary.

Garner: We looked for some rotary converters.

Thelen: Motor generator.

Garner: Yeah, motor generator.

Thelen: And it's my impression the museum wasn't really interested in having 500 pounds of high-speed rotating machinery sitting around as an insurance or personnel risk. So this meant going electronic. And so our first converter didn't quite have the power to do it. So these guys would push the button to start it up and you hear these clunking noises. And it might or might not come up.

And then you began to wonder what in the world is going on with the power supplies with these big relays going clunkety clunk. And for some reason, I decided I didn't want to learn the circuit set here. But I can do linear power supplies. Any high school kid can fix linears.

And so the game plan though on these any power supply's capacitors. And they're about the size of a drinking cup, really. And over the years, the chemistry in them, if you don't exercise them, reverts back to a less energetic form. When you make an electrolytic capacitor, when you start it up, you have to slowly apply voltage to it. And it forms a surface, chemical surface, on these aluminum foil pieces.

Garner: An oxide layer. An oxide layer.

Thelen: It's an oxide layer that forms.

Garner: It keeps it from shorting itself.

Thelen: Keeps it from shorting.

Garner: Otherwise you might have fire and flames.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 29 of 38

1041 Restoration Oral History Panel

Thelen: Otherwise you get very high currents and disaster.

Garner: Fire and flames.

Thelen: Fire and flames. So the PDP-11 people--

Garner: PDP-1.

Thelen: PDP-1 people had done a voltage increase regime, very complicated, da, da, da. And I didn't think we needed to do that. So basically what we did was to put a little electric light bulb in series with the capacitors.

And as the light bulb would keep too much current from flowing into the capacitor and making too much heat. And you could tell that this thing was working in the sense that an unformed capacitor, the little light bulb would be bright and it'd slowly dim and dim and dim as the chemistry oxide on the aluminum foil business got formed. So we did that, gee, I don't know how many capacitors.

Garner: You did it for a year about, I think.

Palmer: Well, we pulled out every one of the power supplies in the tape drives and we gave them to you. And then you had to reform them. And there's another whole bank of capacitors on the bottom of the tape drive. And you reformed all of those for us.

Thelen Yeah. Reformed them all.

Garner And I think only one was un-reformable?

Thelen One had leaked in its previous life and had been dried up. And the internal resistance was just wrong.

Garner: And I think you would agree, Frank and Allen, that power supplies have been pretty solid all these years.

Palmer: Yeah. No problems with them.

Thelen: Yeah. So once these caps are reformed, they work great.

Garner: Yeah. So originally our first 60-hertz to 50-hertz converter was a freebie that Dave Bennett was able to get to us. Then the compliance lab at IBM San Jose was shut down and we had to bid on the more capable unit, the one we have now.

And I unfortunately had to guess what it might cost. Pacific Power is their maker. They still make them actually. So with that unit, we had adequate power. So during that time, during kind of the first three years, Ron Williams on the 1401 was finding a bad SMS card, one or two bad SMS cards, every week. In the end after three years he found about 130 bad cards. And we found a lot of bad cards in the tape drives too, I suspect.

Palmer: Yeah. Yeah.

Garner: I mean, the problem with the German system is it was stored in a garage outdoors for 30 years. Germany is moist. So the transistor cans have iron in them and the leads have iron, they rusted!!

Thelen: The leads and transistors and diodes are iron. Well, the reason they're iron is that iron and glass form a good bond. And you don't have to worry too much about it. But iron rusts. So a lot of the transistors, transistors three-legged thing, one of the legs would be corroded or partly corroded.

Garner: Yeah, you think transistors are solid state, right? So we would think that would be the most reliable part. But now we have rusted leads.

Thelen: The diodes were the same problem. Diodes have iron into a glass thing. And you'd find a diode and the rust had gotten in. Rust was bigger than iron and it cracked the glass in the diode and actually break it off.

Garner: And we would see, you might remember this, Ed, as well as I do, you would see transistors with IV curves that had loops in them, that the behavior would change over time. And you'd have a memory and some of our gates would oscillate and do other weird things.

King: We just had one Monday that would just start an oscillation after it had been powered. You'd use it and once you've used it for a few seconds, once you've got two hours of warm up, then every time you use it for more than a few seconds or less, somewhere in there, it would just start this weird oscillation.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 30 of 38

1041 Restoration Oral History Panel

And so it looked like data was coming in from the tape drives at such a horrible rate that we couldn't handle it and give it all kinds of errors and hang up. We just got that fixed yesterday. We've been working on that for-- that thing's been leaking for months.

Garner: Yeah, Ron would find the loopy transistors and what should have been a flip-flop, they'd turn into a multi vibrator. Shorts. We had this thing called the Overlap feature, remember, in the German system.

Thelen: Oh, we should have Ron Williams here.

Garner: To talk about it. It tied many inputs together. So failure domains could propagate much further through the long circuit paths.

Thelen: And they're talking about a system overlap. And it turns out that you can have many inputs that any one of these transistors fire to make an OR gate. And trying to find out where you're getting the signal to this what they call a wired OR, drive you nuts.

Garner: Basically after three years I was worried that we had just purchased 10,000 rusted transistors.

King: Well, we still have those transistors that are failing that are in the process of failing. They got these little things in them and they're growing and will continue to do that. But it's almost manageable now.

Garner: Well, we have in an air-controlled environment. It was in an uncontrolled environment before.

King: It's already was started. In other words, the rust was already there.

Garner: So now what about the printer? It seemed to me the German printer came right up.

King: No, no, no. The chain had bonds. I took it-- we took that down to machine shop in Morgan Hill. And I had them take enough--

Garner: Off the chain?

King: Off the plate for the chain to get rid of the bond. But not to take too much off that we would play around. The ink and the oil where we had bonded together with a type of rubber, these plates, IBM had done that 50 years ago. And it swelled enough to cause waves in the plates and we had to grind those down. And we may have to do that to the [INAUDIBLE].

Palmer: What about the plastic strip between the hammers?

King: Oh yeah, this plastic guy at Morgan Hill said-- I said we've got to have something like this. And I could just show him pieces because every time he touched it, it would fall apart into little more pieces. And he said, well, Kapton is the strongest plastic there is. And he just gave me a sheet.

Garner: I think I got you the Kapton.

King: Huh?

Garner: I got you to sheet of Kapton from IBM Almaden, I think.

Palmer: But it is amazing when you think about that the machine basically, let's give it 1960, so we're talking about 50 years ago when we started on it over there. Considering how less than well cared for it was, particularly the German machine was, it has come back really, really well. It has some problems and we're always going to have problems with a transistor that's going to go a little flaky.

But really it's still a great machine. It's a joy to work on and to see it go. And to see that thing move 600 cards or print paper so it spits out over there. I get the enjoyment when visitors come through. And when the docents start to talking about the power they have in their iPhone versus what they're looking at, so that they can get some kind of an understanding or appreciation for how the technology has changed within a single lifetime. And so it's still a great machine.

Garner: So yeah Frank, you talk to visitors a lot when they come in. What kind of questions do they typically have?

King: Once they realize that the 1401 was really the first time businesses had moved to a magnetic media away from cards. And I just think that's the most amazing thing in the world, the length of time we moved from roughly 1928 to now until the early '60s. Everything was in cards. All the data in the world was in cards.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 31 of 38

1041 Restoration Oral History Panel

Garner: That's what you appreciate about the 1401.

King: I think that's their look too. Because when they start punching the card, that's the first time most of them realize: “Wait a minute, this is a record and they've put it in the computer!”computer!” I think that's when they really realize what the 1401 did to the world.

Garner: But do you find that many people have never seen a punch card before?

King: They've heard about it. And you've got to be pretty old to remember getting a light bill in an IBM punch card and to writing the number amount you paid and send it back and have that card be the actual thing that's processed.

Palmer: I think it's also when the docents speak to them about how much memory is in it. And wait, this is thousands [of characters], this is not megs.

Garner: 16,000 characters.

Palmer: And that's the biggest one. And the smallest was 1400 [characters of main memory]. And so again, it's appreciation for what it was. But I think it also helps them. I see some of them, they'll look at their cell phones and they'll try to even just start to comprehend what it is they hold in their hands.

Garner: Well, when we remind them how much it costs as well, right? In today's dollars, one of our 1401 systems would cost $2.5 million. So Bill, do you have any feelings of interacting with the visitors? You talk to them sometimes.

Flora: Well I mean, the ones that somehow I get a chance to talk with are people who were either operators, primarily operators. And they know how to put the cards in the reader. They know how to put paper in the printer. And it just makes them feel good that they are familiar with this. And it makes them nostalgia, I guess.

Palmer: They used it in school, a lot of them. The older ones that come through.

King: A lot of them programmed in college. Maybe not on a 1401, but in 360 you still wrote your programs on a key punch.

Garner: What about kids though? I see a lot of kids, kind of wide-open-eyed kids.

Flora: Yeah, but that's, I think, because they've had the opportunity to punch some holes in the card and then they see their name get printed. They're very enthused. I don't know how long it sticks with them.

Garner: Do you guys remember that night where we had 100 people in the lobby and we had them through the 1401 exhibit? Where you here for that? You were here that night??

King: Yeah, everybody had to punch up one card. We had key punches in the hallways and everything.

Garner: There was a lot of enthusiasm this old stuff today. Kind of the steampunk movement. And George, do you have something?

Ahearn: I have a comment about something that's kind of depressing and it's that when we come up with a problem on these machines now, it takes heck of a long time to solve. And it makes me realize that the early development was very primitive.

Like comparing what happened back then in the '50s with today's world, you wouldn't do a machine, a system like this without having a software model of it and doing design verification before you committed the thing to silicon or whatever it is. And the other thing you wouldn't do is you wouldn't put a product out with such deficiencies in diagnostics.

You would like to have diagnostics that could point to a failing replaceable unit. We don't have anything like that. And back in the days when it was being developed, we had a group that knew all about the 1402 and its attachment. And another one about the 1403.

And another one about the memory and another one about the clock and another one about the Edit command. And when you ran into that problem, you'd turn to that individual. And if it meant suspending what you were doing until the next shift started when that person was there, they would come in and address it. But now we have the general thing. Everybody is expected to be able to solve whatever comes up. And it's taking altogether too long.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 32 of 38

1041 Restoration Oral History Panel

King: Well, one of the things we did is we work only a Wednesday or come in once a week.

Garner: Monday now too.

King: We work a few hours a day. The traffic is so bad you dread getting on the road to come up here. I don't know about you guys, but I come from Morgan Hill, which is a job in itself, just coming up here and going home. You cut loose. And we've done pretty much to get back to it to now maybe solve a problem.

Garner: We have Monday sessions now too.

King: Yes, but it would've only taken-- it would probably take, and I would say maybe that one tail problem would have probably taken up to two or three days. It could have been that we've got problems that severe, particularly if you don't have Ron Williams working on it.

Because he was really unbelievably good at diagnostics. Or is. And working on that thing. But you let it go and you think about something else and do all this stuff. But in IBM world, you were going to work 12 to 24 hours straight on that problem until you've got fixed. And people were going to relieve you and you were going to connect with them.

Garner: Did you ever have to swap out a whole system?

Palmer: No. Never. To make you feel a little bit better about it, in actuality when it was in the field, it was not as you're seeing us work on it here today. First of all, you were very, very well trained. 1401 school was 10 weeks, 10 or 14 weeks. You worked on it.

It's not that an individual machine at a customer's was bad that you had a lot of hands on it. But you had a lot of the machines. And if you had a problem that was stumping you and stumping meant because it was down, you didn't wait too long. In my branch after half an hour you put a call in. Another guy would come over and you'd put your heads together.

So because you worked on it a lot, because you were really well trained, because you had a buddy or a tech specialist in your office, or I worked on a machine one time where they flew the specialist up from Saint Louis. He walked in and he said to me, well, what's the problem?

I started explaining. I say, well here's where I am on it. He said, [INAUDIBLE] on a Saturday. He said let's start from here. He looked at me and he said, you've made a mistake or else I wouldn't be here. He looked back at me about six points on it. A little, a wavy line, put a new card in, he got on his plane, went back.

So, really, from a CE who worked on your machine in the '60s from the customer's point of view, the only reason we sold 15,000 plus of them [1401s] was they were damn good, solid running machines. You guys did a hell of a job with the technology at the time. Yeah, you couldn't do all the modeling you just said, but it would have taken you another five years to get it out before the committee agreed on it.

Garner: Yes, some of you guys were conserve enough in your design that there weren't a lot of speed margin issues and stuff. I always wondered how you did that.

King: Well, most of the problems we have are deterioration.

Garner: Today.

King: Today, because that Connecticut machine, even though it was embedded in rust for a few years, wasn't 20 years, but it was embedded.

Garner: It came up much quicker. It came up in six months.

King: Well it comes up and it stays up.

Garner: Actually it came up though in six months and it only had 26 bad SMS cards. So he [its original owner, Buzz Bellefleur] had even maintained up until 1992.

King: And it isn't going bad at the same rate the German machine is going bad.

Palmer: You guys did a hell of a job. Unbelievable. Just to think of the leap you made. I mean, that memory storage, the program that went with, the fact that a little shoe company in the north side of Milwaukee could afford one of those with a couple of tape drives over there and printing out everything else like that. It changed the computing world. It changed the business world. And you did it at a price.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 33 of 38

1041 Restoration Oral History Panel

Garner: The team knew they were doing that too, which is remarkable. Fran Underwood knew this was going to have that impact on the world, as I recall. Did you remember that, George? George, right? Fran knew this machine was going to have this impact on the world.

Ahearn: Yeah. Back to what Allen said. I appreciate what you said. Because the reasons you gave explains why it went so well. That is, people worked on it all the time. People knew what it was. People were well trained. People had colleagues who you would call in after half an hour. And if you couldn't get it, someone would come along. And then there was another time limit where you'd go to the region and bring someone.

Palmer: Yes, absolutely.

Ahearn: Those are all the reasons why it went well. We don't have benefit of that now. So we have to do this we can with limited knowledge.

Garner: Yeah, and younger people are beginning to learn how the machine works. Guy Fedorkow and Ken Shirriff are now writing a large theory of operations manual for the 1401. You guys haven't had a chance to review it yet. But I also wanted to ask Iggy, you came on what, a year ago?

Menendez: About eight months ago.

Garner: What has been your feeling since you've been on the team? About the machine, about the restoration project?

Menendez: First of all, I love it. It's like restoring a Model A Ford. So I really enjoy it after being away. I retired and went away for 15 years. And coming back is just like a hint of life again. It's incredible. It's just absolutely, I look forward every week to come here and work on this and work with these guys.

Thelen: These guys? Boy, you got no taste.

Menendez: You too.

Thelen: OK, thank you.

Menendez: But it's really incredible. I love it. It's great. Frank invited me to come, where we're meeting another group on Fridays. And he kind of mentioned in passing. I had talked to Chuck. Chuck Kantmann also good friend of mine from the beginning. And when I was retired in Mexico, I had some conversation with them about the museum.

Garner: Oh, you knew he was up here at that time?

Menendez: Oh yeah. And then when Frank mentioned it again at that meeting, I said OK, sure, I'll go. And I've been coming back ever since.

Palmer: My golden prayer is that you do not find another 1401 and buy it and bring it in.

Garner: Well, we found that 1440, remember, and we donated it to the Center For Technology Innovation in Binghamton. And that then attracted designers of 1403 and people in Endicott to restore it. George, you're going to visit there.

Ahearn: I got to talk to you about that.

Garner: OK, later, not during the interview. So anyways, that was an excess machine that we actually sent to Endicott. And that's really a good place to be restoring these machines. And then subsequently we got a 1403 and we got that 2501 card reader. And there's an opportunity they actually might get Paul's 1401 system.

King: Well they've got print positions on that thing working. That's amazing.

Garner: I'm also saying that there's an opportunity that they may get Paul Pierce's 1401.

Palmer: Endicott you're talking about?

Garner: Endicott. Binghamton is actually where it is.

Palmer: Oh, Binghamton, yeah.

Garner: So Ed, tell me, you've been the webmaster. You've been called the captain by--

CHM Ref: X7542.2016               2015 Computer History Museum               Page 34 of 38

1041 Restoration Oral History Panel

Palmer: He's our Boswell.

Garner: Yeah. You've been called the Garrison Keillor of computing because you maintain all these computing websites. You like to record everything for posterity, so talk about that a little bit.

Thelen: Well, when I was a kid I liked history. The thought is that if you can learn from the mistakes of your predecessors, you might not do the same dumb thing again.

Garner: That's optimistic.

Thelen: Yes, it is pretty optimistic. People want to blunder their own way through. But that's sort of the driving force is if you ignore the past, you might not do the same dumb thing again.

Garner: Technology is so different today.

Thelen: Oh yes, yes.

Garner: But don't you get some joy out of just uncovering and presenting the old history? I mean, you have all these history sites where you've recorded the stories. People have written hundreds of stories. John Gartner is one of them. He's featured under stores on your web site You guys need to write some more stories. You've recorded a variety of machine histories, not just IBM and the 1401. So I think you enjoy the hunt and the exposure, I think, right?

Thelen: That could be. There's probably some selfishness in there too.

Garner: Well, you like sharing it though. Because you always seem to have it in your blood, let's get this on the website.

Palmer: If it wasn't for Ed, there really wouldn’t be that history that he kept every single week with pictures.

Garner: Yeah, Ed, you kept, if anyone goes and looks on the site, you kept a daily log of what happened every day with pictures. You recorded the whole history of the restoration project!!

King: I've had compliments all the time from people who started reading some of Ed's stuff. An old manager of mine ran across it and he just couldn't get enough. I mean, and practically everybody he worked with back in our office. They all seem to know all about what we're doing out here.

Ahearn: Was that Hibbs? Hibbs? Don Hibbs?

King: No, what Ed wrote.

Ahearn: I mean the manager you talked to.

King: Oh, no, that was Clete [INAUDIBLE] back in Washington, DC. He started reading it and couldn't quit. He said he was enjoying.

Garner: Ed's writing style is--

King: Is so good.

Garner: Very affable. I like to call him the Garrison Keillor of computing. It's pretty funny. And it's insightful. And I think he should write a book, but that's a different topic.

Thelen: This is my book.

Garner: Oh, OK. So on another topic, someone who can't be here with us today because he passed away last year was Bob Erickson. I think Bob touched us all very closely. He was what, 93 when he passed away?

Thelen: I believe so. 92 or 3.

Garner: And he from memory restored the unit record equipment, each with 50,000 components in them. You guys have any stories about Bob you'd like to share?

Palmer: I think that the picture that is the epitome of Bob Erickson is rebuilding the memory. Now that's something you'd never do in the field.

Garner: A broken core memory.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 35 of 38

1041 Restoration Oral History Panel

Palmer: Yeah. He was stringing the wire. That's something you'd have never done in the field. If that actually happened in the field, you'd have pulled out the memory and just put a new one in. And to see him sit down there and rebuild, I think he was working on the 557, the interpreter.

Garner: Yes.

Palmer: Just from memory and spreading all those pieces out very neatly, very clearly. What a mind. What a mind.

Garner: Yeah, he was one of the most smartest people I know at that age.

King: He wanted to rebuild the 513 when I had perfect shape 519 from Portland. Was almost brand new. And he wanted to rebuild a 513, which was an old motor-generator thing with curved legs on it. And when we were getting ready to put the punching back in, getting the punching it ready to put it back in, we had data over in stored.

And I said, “We better put these pins back in first because we won't be able to get this thing in unless we get the pin in first.” And he said, “Oh no, IBM cut a hole in the side of that thing so you could get the pin in later.” And I just made the comment, “How the hell would you know that?” And he said, “Oh, I worked on these things before the war. Now he's talking about World War II!”!”

Garner: He had a perfect memory.

King: I mean, and he could remember that there was a hole on the side of that thing. I couldn't remember what happened yesterday.

Garner: And he restored the 077 with you as well.

King: Yes. He spent endless time with that worn out German 50-cycle piece of junk. Oh yes, it works.

Palmer: He was very, very interesting, very gentle person. Very humble. Didn't tell many stories of his times. But it was interesting how he, on his career, and how he was designing machines and everything else. And that's what got him hired and then moved to Washington to I think Endicott where he was working.

Garner: Well one story I remember was when Gene Amdhal came to visit, he said, “Hi Gene.” Because he had designed the tube module for Gene Amdhal on the 709. He designed the index feature on the 709.

King: I think we have a picture of that.

Garner: Yes, we do. He was at Los Alamos during the atomic bomb testing. So he told us a lot of stories about Los Alamos.

King: And he was involved in code breaking equipment during World War II in Australia.

Palmer: Yeah, he was in Australia because that's where his wife was from. He was in the South Pacific in Australia and met his wife. Very interesting gentleman. One of the group.

Garner: All righty.

Palmer: The project has brought a lot of interesting people.

Garner: So yeah, I'd like each of you to think about its future just a little bit. And I hope it's a little positive. But any thoughts about where to go from here? Iggy, do you want to start and maybe we'll just go across? What do you think it's going to be like 10 years from now on this project?

Menendez: Well, I would personally love to see some of the following machines come to the museum, the ones that I'm familiar with more than the 1401. Some of the early 360s and some of the funny machines like the noodle picker. Like the 1288 optical scanner. Those were really fabulous machines that were [INAUDIBLE].

Garner: You want to keep getting stuff put together. I appreciate all the work you're doing on the tape drives, by the way. Because as Allen will tell us, they can be temperamental.

Menendez: Thank you.

Garner: Ed, what do you think about the future? 10 years from now, what do you think?

CHM Ref: X7542.2016               2015 Computer History Museum               Page 36 of 38

1041 Restoration Oral History Panel

Thelen: Well, I hope that we get enough information on the website to help people that might want to come in. Younger people under 75 to help restore these, keep these things running. The logic of the Edit command. Or if they have a memory-core driver failure. If we get enough information so that they can learn what these people learned. Or at least a good clue. Then maybe this thing can last 10, 20, until the bearings wear out.

Garner: Can always buy new bearings.

Thelen: OK.


Palmer: What I would like to see at the end of the 10 years would be that during this period of time, the museum really took it into their DNA of a true interest in maintaining it. Such that when John Hollar had different people come in during his next years planning of the five-year plan, someone on the staff other than us would be responsible for putting together the plan to maintain the 1401 system.

So if you went up to John Hollar today and you said::“What is your plan to keep this system running 25 years from now?” My answer is he would say:: “I'm really busy. I gotta go to the meeting.” So what would I like to see? I'd like to see the museum get dead serious about maintaining this.

King: I'll have to agree with Allen on that. Because right now we are so technically oriented and to the museum we're the only people who can make it run. But that's not really true. They need to put some effort, I think, into the colleges and the high schools and things to get young people to come in here and be apprenticed with us.

Because they're smart today. They can fix these things. They can learn the logic so fast. I mean, at their young age they can learn the logic so fast. And we need to put somebody in the museum. Like Allen says, that's responsible. Some team, some group that has the responsibility for keeping things running. And I mean go to college and give them college credit for coming in here and working as our apprenticed.

Palmer: How they did it is up to them, but they need to do it.

King: Right. We need to do it.

Garner: OK. George?

Ahearn: I would like to see a disk drive get attached.

King: Yes. Absolutely. Yes.

Garner: We’ll work on it.

King: Well we lost our disk drive to the 1440 people, right?

Garner: We did, yeah, we did.

Palmer: Did we give it to them? I love that 1311.

King: Well the 1311 is how we wound up with the 1440. Because II kind of didn't want to steal the 1311, because that was the basis of that 1440.

Palmer: My jaw dropped when they announced that machine the first time.

Garner: OK, George, anything else?

Ahearn: Back to the business of younger people, I would expect that a younger person is concerned with their resume. And putting this kind of thing on their resume doesn't do them much good in the job market.

King: Depends on what you're going for.

Garner: That's an interesting observation. In Binghamton, they've got young kids at technology schools who built the PC-to-1403-N1 interface. And it does look good because they used a modern [PIC] controller and they proved they could do something with their knowledge. So maybe there's some way that you can combine maybe an interface to a disk drive, for instance, would be necessary.

Palmer: I'll make you the argument that [INAUDIBLE]. Teaching programming, and a 1401 is a good machine to program on, is an excellent training of the mind to solve a problem. Because once one of the biggest mistakes the early programmers made, when you do a compare, is it high, is it low? But they forget, is it equal? And so I think it does work on their resume.

CHM Ref: X7542.2016               2015 Computer History Museum               Page 37 of 38

1041 Restoration Oral History Panel

Garner: Yeah, Steve Wozniak was here. Steve Wozniak thought this was the coolest exhibit in the museum because it could teach kids all the way from transistors to programming in an obvious way.

Palmer: And they can touch it.

Garner: Touch it, yeah.

King: It's down to that level. I mean, we can fix a [INAUDIBLE] on another transistor and stuff like that.

Garner: For me it's been a great joy to see people like Ken Shirriff, the Google guy, the Google map guy, taking extreme interest in our 1401 programming array tracing algorithm and a bit mine coining algorithm. All these things. So Bill, what are your thoughts for 10 years from now?

Flora: Well, I mean, I don't know what to say here. Because we obviously do not work for the Computer History Museum. And somehow the youngsters are sharp, I understand that, but they're not committed too. So I think the Computer History Museum needs to step up to putting some kind of a plan for the future.

Garner: Common theme. Any last thoughts? You guys have been great in this interview. Any last thoughts?

Palmer: Thank you.

Garner: All right.

Palmer: Thank you for the opportunity to spend 10 years with friends. 12 years.

Garner: I think I get the better part of the deal. It's been so much fun working with you guys.

King: I think it's been a great pleasure to work with all these guys.

Palmer: A lot of us worked together in the FE.

Garner: Well, you guys have been a joy to work with the last 10, 12 years. So I look forward to the next 10, 12 years.

King: That means we're not giving him a hard enough time.

Garner: Oh yes you are.

Palmer: Does it mean we're going to get a raise in the next 10 years?

King: He doubles our salary about every month.

Palmer: You know, he used to-- wait, wait. What do I want in the next 10 years? There used to be a guy who came in on Wednesday with bagels, two different cream cheese. Yeah. And there was a time when the Computer History Museum had a room for us to eat lunch. It had cold sodas in the refrigerator.

King: Oh yeah.

Palmer: Who was the bagel guy?

Garner: That was me. OK, all right, thanks, you guys!!


CHM Ref: X7542.2016               2015 Computer History Museum               Page 38 of 38