Monday, March 15, 2010

Tools - 1973

In 1973 "State of the Art" tools were different than today's. Above is a picture of the main tool we were provided with - a flowchart template. This is the actual official IBM one I was given and that I still have it is telling.

We used this tool for two purposes. First, during class we used it to create a diagram of our logic and then wrote COBOL code from the diagram. For most of us this use ended when the training class did. Second, this template was used for creating work flow diagrams that accompanied job requests to the actual operators in the Computer Center.

When I arrived in Management Financial Services we used magnetic tapes for data storage. At that time programmers would actually duplicate production tapes or create tapes with partial data sets from the original production tapes and use those for testing their programs. Depending on the size of the project we had stacks of tapes on and around our desks. The programmer would create the Job Control Language (JCL) program on a series of punch cards, put those in front of the punch cards containing the program code to create a deck and use the template to draw a diagram for the computer operator to show him or her what tapes would be called for, when they would be called and how they were to be mounted on drives for the program. The diagram was drawn on a special form (white, 8 1/2 x 11 with identifying info in boxes on the top) which was wrapped around the punch card deck and attached to the tapes for the job with elastic bands. These were placed on tables located around the department and picked up at various times by people who came around with carts. Finished jobs were delivered with output the same way. It was not uncommon for a programmer to go directly to the Computer Center when he or she had an important ("priority") job.

When the Tower was built the bottom floor was designed as the Computer Center and had ceiling to floor windows so the company could show off its million dollar computer hardware. After a series of bombings at some universities in the early 70's, the windows were replaced with thick concrete walls and later the entire Computer Center was moved underground.

Tuesday, February 23, 2010

Training Classes

In early 1973 I returned to Hartford from my working visit at the Boston Regional Office. The Education Center is housed in a former insurance company building on Gogswell Street that Hartford purchased after the original occupant went out of business. I was part of a group of 10-12 people, all of whom had just spent time in various Regional Offices around the country.
We were housed together in top floor apartments rented by The Hartford in a nearby building and got to know each other pretty well as we spent our days in classes together and nights in various apartments - not unlike college days.
As training neared the end (3 months?) we had a series of speakers come into our classroom and talk about their various offices/departments. From this, our visits to the Regional Offices and some tests we took, we were supposed to pick areas of interest and if there was interest in us from the departments, unofficial interview visits would result. As I hadn't seen anything at the Regional Office that I was really interested in, I ended up talking with two Home Office departments - Quantitative Methods, where they were trying to build a model of the company and the economy so they could predict rates, profits and losses, and Systems, where they were hiring twenty somethings like crazy into programming positions.
The selection process was interesting. Each of us candidates received a piece of paper with three numbered lines on it. We were to fill in each line with our first, second and third choice. Then the departments would get to select the candidates they wanted. Hopefully the department you wanted also wanted you. In my case I decided I didn't want to work in FORTRAN on an economic model so I selected Systems on line 1 and left the other two blank. Confusion ensued. I found out later that Quantitative Methods really wanted me, but because I hadn't even made them a choice they couldn't have me and I ended up going to Systems where I was an acceptable but not top choice. Even later I found out they changed the process so in the future everyone had to fill out all three lines...
After being selected for Systems it was back to training again to learn COBOL for another 8-12 weeks in the Education Center. My Management Training classmates were scattered throughout the country and I never saw any of them again. The COBOL training was basically a beauty contest with job assignments being based on final class rank. All of us except one person finished the class and were assigned to various divisions within the Systems department. I was assigned to Management Financial Reporting Services in a work group writing and maintaining programs to track losses. The Systems department was growing rapidly and I ended up on the 5th floor of the Tower with 100 other twenty somethings. The oldest people on the floor were the supervisors and the officers.

Monday, January 25, 2010

Hartford Insurance Group

In Fall 1972 Eric Johnson, a high school friend of mine I had kept in contact with, had finished his Army service and was working at ITT-Hartford (Hartford Insurance Group) and suggested I apply there so he could get the $50 bounty for referring me. I believe he knew someone in Personnel so I was invited in for an interview and selected for the Management Training Program. At the time the program was a short cut for moving into positions of responsibility at the various regional offices of the Hartford. I found out later there was some friction between the graduates of the program and some of the old-timers who had "worked their way up" to their positions.

I was assigned to the Boston Regional office which was located at Government Center. I got lucky and found a small 1 room apartment on Beacon Street and took the subway into work each day. I got even luckier with my car which I parked over by the Fenway on Sunday nights and walked back to on Saturday mornings. In three months all it suffered was a broken tail light. I was moved through the various departments of a Regional office and in three months went back to Hartford for classroom work armed with my knowledge of regional office operations and an opinion that there wasn't anything I really wanted to do at a Regional office...

Monday, November 2, 2009

GE 635

Like most college graduates in 1969 I had two choices; go to grad school or go to the army. I chose grad school and in September entered the MBA program at the Amos Tuck School at Dartmouth.
At the time Tuck was in three buildings at the end of a road in a corner of the Dartmouth campus. Our computing facilities consisted of a small number of ASR 33 teletype machines in various locations around the school which were all wired to the GE 635 "mainframe" at the Kiewit Computation Center up the road behind Baker Library. The GE 635 ran the Dartmouth Time Sharing System (DTSS). The teletype machines had paper tape readers for loading your data or custom programs and printed all input and output on the rolls of paper fed through the back. Most of the programs we ran had been written in BASIC expressly for financial data or statistical manipulation.

The history leading up to 1969 is interesting and I will relate some of it here, but a complete retelling is available on the Dartmouth Computing site.
1962 - Dartmouth mathematicians John Kemeny and Thomas Kurtz proposed building a College computation center and chose to implement a "time-sharing" system based upon a concept first realized on a small Digital Computer Corporation PDP-1 computer by a team of scientists from MIT and Bolt, Baranek, and Newman, Inc. They request a NSF grant to fund the project.
1964 - The NSF granted $500,000 to Dartmouth for the development of a time-sharing system and the computer language BASIC. A GE 225 computer, plus software, carring a price tag of $800,000 becomes operational in February.
1966 - General Electric (GE) renamed DTSS as the Mark-I and used it to build the largest commercial time-sharing system. Based on the success of this work, GE and Dartmouth College embarked on a project to put DTSS on the newer and larger GE-635 computer capable of handling 200 simultaneous users which becomes operational in 1967.
1969 - GE gave Dartmouth title to the $25 million GE-635 computer that had been jointly operated for three years, now serving 300 terminals at Dartmounth and throughout New England. The College and General Electric also announced a new three-year partnership in "cooperative work in computer technology."

In 1970 GE sold its computer business to Honeywell who later moved away from time-sharing. At the time, the sales event did not make a major impression on most Dartmouth students as there were other events happening at the time in Laos, Cambodia and Kent State.
I enjoyed using the terminals, running the programs and playing a little with BASIC, but still failed to envision the future of these machines...

Monday, October 26, 2009

State of the Art - 1969

When I graduated from college in 1969 the Computing Industry consisted of IBM and a group of (much) smaller competitors. These competitors were called the "7 Dwarfs" and eventually morphed into the "Bunch". They were Burroughs, Sperry Rand (eventually to become Unisys), NCR (National Cash Register), Control Data, Honeywell, RCA and GE. Most of them competed with IBM only in niche or specialized markets.
I looked up the 1969 Fortune 500 to get a feel for the relative size at the time of industry in general and the computer manufacturers in particular. What I found was surprising. IBM was the 6th largest firm. Of the top 10, three were auto (GM, Ford, Chrysler), four were oil (Exxon Mobil), Mobil, Texaco, Gulf), one was steel (U.S. Steel) and one was diversified (GE).
IBM was 6th with sales of $6.888 billion. GE was 4th with $8.381. RCA was 20th with $3.106. Sperry was 56th with $1.562. Honeywell was 72nd with $1.281. NCR was 86th with $1.102. Burroughs was 156th with $650 million and Control Data was 221st with $438 million. Overall industry was a lot smaller in 1969 than it is now. The largest firm was GM at $22.755 billion sales, the 10th largest firm was U.S. Steel with $4.536, the first firm over $1 Billion in sales was Olin with $1.002 in 104th. The last place on the list was held by Briggs & Stratton with $143.7 million in sales. Of course this was before business college was popular. A significant number of my classmates were sons of owners who had been sent off to learn "business".

IBM 1130

I attended Nichols College as an undergrad from 1965 to 1969. In the Fall of 1968 when I returned as a senior, there was a new IBM 1130 installed in the basement of Conrad Hall, a small building that contained the President's office.
Later that year, I think it was during second semester, I took the first and only course offered to students on this machine at the time - in FORTRAN programming. We wrote the code for our assignments on a special paper form that was green and white and had 80 boxes per line which corresponded to the 80 columns available on punch cards. I don't remember actually punching the cards so I think we left the sheets in a basket and came back the next day and the cards were wrapped inside the coding sheets. We then placed the cards in another basket and the jobs were run overnight. Ah, batch processing...
I learned FORTRAN this way but also that the machine was unforgiving. Any mistake, no matter how minor would cause the program to end abnormally (abend) and since we were limited to one run per night, we ended up learning to be very careful about error checking our code.
I was mildly interested in the computer and coding, but not enough to see the potential.

Thanks to my classmate, Ken Burrill, who wrote an article in the Nichols College magazine (see page 34) that refreshed my memories about all of this.

Friday, October 16, 2009

Binary State

Some days I feel like this:

And some days I feel like this: