Return to More Liberator
A partial transcription of Edward Terry's visit
Recorded Wednesday June 27th 2012, after:
- a tour and demo in 1401 Restoration Room
- the scheduled Noon Tour of R|EVOLUTION
- a cheese quesadillas and rye/sliced meat lunch from the in-house restaurant
while sitting around a table on the NW patio, under an umbrella ;-))
The participants are - Edward Terry, his sons Dale, Dave, Darren, daughter-in-law Joanna, Bill Worthington (docent), Ed Thelen ( very curious about Liberator ). The recording lasted 54 minutes.
This transcription has somewhat smoothed out the pauses, partial words, ... of normal informal speech. Some asides and family stories are not present.
Special attention is paid to exact wording of Edward Terry's comments about the Honeywell Liberator software for converting IBM 1401 object decks to Honeywell H-200 series object decks.
00:11 ... the Sacramento IBM SE manager wanted me to come in and talk with him, so I did, I went in.
The first thing he asked me "Jim tells me that you are interested in Honeywell. How far along is that?"
I told him the truth. I said that I have accepted an invitation for an interview with them, but I haven't interviewed with them yet.
He told me "I'm sorry to hear that."
This was my first experience with IBM ethics.
He said "I'm sorry I can't talk with you because of that."
So instead of getting the opportunity with IBM ...
01:30 So there is that fork in road again,
I could have gone to work for IBM, if I had kept my mouth shut,
but as it was, I accepted a position with Honeywell.
01:40 The first two or three weeks I just hung around the office.
01:56 But the classes weren't being scheduled for a month or so back in Wellesley Hills, near Boston, that's where the education center and Home office Honeywell Information Division - 02:22 Sent to account down the peninsula [SF??].
They were trying to sell, Westinghouse, write a 1401 program, so I did, trying to keep me busy
02:58 Went down for about a week, wrote a couple of programs for 'em 03:08 Somebody from Honeywell go down write 1401 programs for Westinghouse ??
Somebody to demonstrate to prospective Honeywell customer IBM 1401 knowledge
03:37 If / when Honeywell 1800 replaces IBM 1401 03:51 Then I started my training back in Wellesley Hills I went ot ??? three ??? schools - spring of 1964 04:07 First 2 weeks were on the H-200 hardware and programming 04:20 And then there was a delay of a week , a layover, 04;24 Which I didn't want to do, I wanted to come home. but ??? said, stick around, stick around, I've got you scheduled for the Liberator class.
Nah I want to go home
04;35 He said "Why don't you go down to New York, to the World's Fair, and I'll OK the expenses" - 04:50 So that is what I really wanted anyway, so ... 05:20 So anyway I came back for the Liberator class
That was just a week's worth of instruction.
05:40 ... HOW TO run the operation of the Liberator program itself 05:52 but when I got back to the office, they had already sold Singer Sewing Machine over in San Leandro and they wanted to put me to work so they sent me down to the Honeywell Computation Center in LA 06:10 Located on Wilshire Blvd near Figueroa
I spent the summer of 1964 through approx Oct. or Nov. Down there 6 or 7 months -
06:33 I stayed at the Canyon Motel on Figueroa St. just a short walk to the Honeywell Data Center.
They called it the Computational Center.
06:52 and they had an H-200 that you could see from the street, you could be on the sidewalk and you could see all the tape drives and all the equipment.
It was a raised floor too.
07:12 Singer started sending me packets, thick manilla envelopes.
They had a source program and an object deck, test data, and results of the test data.
07:43 The proof of the liberation is the comparing the output of the 1401 to the H-200.
If it is the same output that means liberation.
The liberation is liberation, yes, it is a marketing ploy ...
08:11 and the H-200, that machine could whistle Dixie.
It could run circles around the 1401. Much more powerful design, you know -
The 1401 had fixed I/O areas ... card read area from 1 through 80 - I used those areas for program initialization though -
09:05 Before you fired of an I/O, you could use those areas for house keeping. 09:34 I put all my program initialization in those areas, you are only going to use them one time. .. actually program setup. When memory was so confined, you had to do that -
I think we had only 2 K of memory - we didn't have much to play with -
10:06 When the 1401 came to AeroJet, we had to convert all those 407 panels to 1401 in order to use the new print chain on the 1403, which is a marvelous printer. 10:25 ( from Bill Worthington - did you use the Report Program Generator ? )
I used that and I also used FARGO, FARGO was a load and go program, and you could program up coding sheets, and it would allow you to emulate a 407 and so that's what I did.
Gotta get rid of all those 407 panels, we had maybe two or three hundred, I forget how many,
They were just all over the place.
11:06 ( from Ed Thelen - that was your program library )
Ya had to do something with AeroJet you know - so they gave the ??? reading ??? test to everybody, including the operators in the machine room.
11:33 I was in the analysis section at the time.
I was promoted there.
I was one of three 407 board wirers and they sent me to FARGO class at IBM - it was only one day or so - but I also coded in SPS - the predecessor to AutoCoder.
12:20 We didn't stay with SPS after AutoCoder came in - I really enjoyed the work 12:30 ( Dad - what about Singer ? go back to Singer )
Oh - the procedure - the methodology was that they would send me a packet which consisted of
- a 1401 object deck,
- and a ????,
- and a program listing of the source code,
- and any output from the test.
and they had a big machine room at Singer -
I ??? about 400 1401 programs.
13:30 I didn't meet many of the Singer people - they would sent me packets -
Another time at the Computational Center, they had a full blown H-200
I don't know what the model number was but it had 32 K and it had either 6 or 8 tape drives.
14:16 So they had a full blown H-200, and I wasn't the only Liberator down there -
There were a bunch of other guys, and we were working all hours of the day - the machine room was accessible 24/7 .
And when I went down there, I was living out of the canan ?? Canyon ?? Motel - right out of my suitcase - so I gave up my ...
We were living in a boarding house on Pacific Ave in San Francisco - there is a story behind that too.
15:01 ( something about co-ed ) It was really neat ( some laughter ) so anyway I was on expenses for about 6 months down there ( maybe LA ??? ) And since I was on expenses.
I banked my salary - it was a real sweet deal.
15:30 It was my first experience with an expense account ( So Dad - who was the company you talked about in the past where they wanted proof that you could liberate a 1401 ? )
OH - That was Friden Calculator Company - the salesman was selling 'em and they hadn't quite made up their mind if they wanted to go with Honeywell or not -
16:09 So they sent a couple of programs down to me and I went ahead and converted it for 'em - and sent the results back so that peaked their interest so they had a half dozen people come down to the Computational Center for a demo and I ??? the equipment.
And the salesman came down with 'em and escorted 'em in and introduced them to me and ??? equipment
and I showed 'em one of the results of the the programs they sent me.
16:49 That I had converted - the supervisor of the programming staff had a brief case and he reached in and he said
"I want to see you convert this one."
He was a young salesman, he was only about 5 years older than me - he almost lost his lunch - ( much laughter )
He said that I have to have time to analyze - you know -
And I said "Give me that !!" ( more laughter )
17:31 I went through there, and I saw a couple of tricks in there, so I went in there and patched his deck - then I ran it through the Liberator converter which was a two pass operation. 18:05 The first pass consisted of a deck of approximately 2 inches worth of H-200 object deck -
Every time I did a Liberation, I had to put the 1401 in the middle -
So the first pass would go through and it would read the IBM 1401 object deck, then it would come to the end-card if the IBM object deck,
and it would process and you could see the tapes moving.
18:38 The tape drives were actually moving - I remember that -
I got to thinking of the configuration, and once I recalled the configuration, then oh yeah, those tapes were moving -
19:03 So what ever the logic was, it was causing the tapes to move.
And that told me it was either a compilation or process -
I didn't really understand how the AutoCoder Assembler worked but it was sort of a compilation process too -
( from Bill Worthington - sort of a pre-compile ) Right - this was reading the 1401 deck and putting it in such a fashion to make it easy for the second pass to work on it -
19:45 So it had to put the 1401 object in a certain pattern , a file if you will, that was easy for the second part of the Liberation process to work on.
( from ?? - was that the AutoCoder deck or the object deck? )
Strictly the object deck -
20:10 Well, they gave me the AutoCoder listing in order for me to take a look at the logic used - because there were certain things you had to look out for that would cause problems later on -
So you were either patching the 1401 side or patching the H-200 or ???
20:44 ( from Ed Thelen - you mentioned patching the 1401 side ... )
Yeah - I was trying to recall what I did - I think it had to do with address modification - that was the only thing I could think of -
Its so hazy - but I did do most of the patching on the 1401 side ??? procedure ???
21:14 But I recall those tapes moving ??? deck
so the lower deck itself consisted of phase one H-200 code - that was H-200 octal code - and then - it was an object deck - octal -
The H-200 was an octal machine - there was a story behind me knowing what octal was about.
22:01 I could talk for hours about how I became trained - I think hat would make an interesting discussion - because I don't know if I was typical or not - the first nine years I was a user - I used IBM equipment - I was fresh out of high school ...
19 years old 1st job McGraw-Hill
clerk-typist and IBM abstract reasoning test - test was fun :-)) even though I was a poor student in high school, I had the aptitude ....
24:35 ( a son - "Give me those cards!!" )
Oh yeah - I showed the programming supervisor from Friden Calculators, and I showed him the process -
Fortunately, the systems I worked [ converting with Liberator ] on were card systems - they didn't have any tape - tape conversion was a lot tougher !!!
25:20 I was doing these programs, maybe three or half a dozen a day, my work measure performance was the extreme
( a son " six a day ??" )
It was easy because I already knew the 1401, I knew the tricks,
I don't remember all the instructions, but boy I'll tell you there was one I'll never forget -
that was Branch and Store B Register, because that gave me the ability for structured programming !!
I was doing structured programming before it became a popular buzzword.
26:15 ( from Ed Thelen - "That was the way you could call subroutines with that instruction.)
I call it subroutine coding -
the 1401 was so easy to program because you dream up all the subroutines ahead of time like your always going to print, printing is always going to have error checking, or producing tape.
You are reading tape and you are always error checking,
so I got so I had a library of subroutines that I used - I could put together a program really quick !!
27:01 It was the equivalent of macros, before the macro-extension of AutoCoder came out.
When I left AeroJet, the 1401 was really in it's hay-day, and the AutoCoder assembler did have macro capability.
27:41 ( from Bill Worthington - " IOCS - Input Output Control System ... " )
Yeah - there were some analysts that were good programmers, but when the 1401 came in separated some of the old timers that wrote unit record procedures -
They got left behind because they couldn't pickup program logic -
But a guy like me, I was a board wirer, I could easily adjust to that and since I knew Hollerith - the 80 column card -
I was in operations a number of years. Then I went over to the analyst section.
28:33 ( a son " Dad, how long were you at Honeywell? " )
I was at Honeywell approximately 2.5 years -
I started with Honeywell in February 1964. Kennedy had been shot and I interviewed with Herb Greenberg in December.
( a son " And what made you leave Honeywell ? )
Laughing, "I got too big for my britches."
29:02 ( from Ed Thelen - "The story about the red hot program?" )
OH - I went ahead and checked the listing and he had done and I had to do a little bit of patching the front end so I went ahead and demonstrated the Liberation operation right there before his eyes.
So I was able to convert it within the hour and show him the 1401 print listing comparing with the H-200 listing - and they were identical.
30:01 ( someone - " did you have to patch the deck he supplied?" )
What I liked to do, if I ran into a condition that needed patching,
I liked to patch on the 1401 side.
( someone - " So what did you do, replace cards?" )
I'd just go over to the key punch and put in a branch instruction somewhere that ??? to the patch, then wrap up the patch and do what ever I had to do
and Branch and Store B Register type operation.
30:40 After that demo, Friden became a customer and I did some conversion work for them too.
I was down at the Computation Center all through the late spring of 1964 through October/November - so I was down there about 6-8 months.
Living out of a suit case - when I was separated from your mother.
31:12 Instead of hanging around singles bars, I was camped out at the machine room.
( a son - "You had whole access to the computer room?" )
I had a key to the front door, I could come and go as I pleased.
And I did, I did a lot of work at night, you know - when it was quiet, no one around -
31:34 It was a wonderful period of my life, I absolutely loved it.
If you can love a machine - well I was passionate about in those days -
31:59 ( a son - " What other accounts did you do for Honeywell? " )
Well, I was a Liberator only about nine or ten months.
They sold The [ California ] Department of Education on the H-200, that was in Sacramento,
So the salesman sold - I was part of the post support team -
They had already sold the Department of Education in Sacramento and what they wanted to do was install the H-200 in the county schools.
32:49 They used the state as a means of marketing to the counties and I installed H-200s in three counties,
- Ventura county,
- San Mateo county,
- and Sacramento county.
So I was the SE assigned to that and that was all ??? based 200 work,
It had nothing to do with Liberation, I had already left that behind me.
33:28 My job with that, my assignment, There were two sides to education.
One was actually education which included class room scheduling, grade reporting, those applications. They had to do with actually educating someone .
And there was the other side , teacher payroll and accounting. Those were the business applications and I was assigned to that side.
33:55 I did the requirements analysis for teacher payroll, and I ended up installing teacher payroll systems.
( from Ed Thelen - "Did you use the COBOL language or something like that ?" )
No - the first one was in EasyCoder, that was done for Ventura County. They didn't have a programming preference - they didn't have a programming staff. So I did the programming. I got the teacher payroll running.
34:45 I talked to the people in the office -
A teacher might get a salary and also drive a bus, and get paid by the trip - what is the formula for that.
... I wasn't able to completely convert - it was what I call Ivory Soap 99.44% -
I allowed for some manual intervention.
The people in San Mateo County wanted it in COBAL because they had a small Programmer group - I had to learn COBOL
I was sent to COBOL class in Seattle,
The first thing I wanted was a complicated COBOL program, a COBOL listing,
so they did - a master file update -
you have your master file input, and your transactions coming in, and then you have an updated master file, plus a printout of some kind.
I wanted something with an input and an output.
I learned the PICTURE statement that way.
I actually learned more from that program than I did from class room instruction.
37:42 ( from Bill Worthington - " Your job with Honeywell was very similar to mine. we essentially became an ajunct to the customer staff." )
Exactly, - once I got started rolling in Sacramento I was sent to Diamond Walnut which was one of my accounts in Stockton and I became as an extra programmer for them as they had bought an H-200 and the application was Grower's Accounting for Walnuts.
38:19 ... general rambling by all ... 39:07 After a while I identified more with the customer more than the company I worked for -
I was really into customer support. They really appreciated what you were doing for them.
More general discussion of career paths by Edward - not specific to computer technical details
43:40 I had a chance to compete with the [ IBM ] 360, so I was with RCA for 5 years, and only 2+ years with Honeywell.
... and family interactions ...
53:00 RCA and video composers - replacing Linotype machines etc. - next time -
Time to leave for quick view of Honeywell advertising animals,
exemplified here, and medical appointmenr