Software Development at the Eckert-Mauchly Computer Company Between 1947 and 1955

Iterations – Norberg – Software Development at EMCC 

 3 

 

Very early in the history of EMCC, John Mauchly assumed responsibility 

for programming, coding, and applications for the company’s planned 

computer systems. His early interaction with representatives of the Census 

Bureau in 1944 and 1945, and discussion with people interested in 

statistics, weather prediction, and various business problems in 1945 and 

1946 focused his attention on the need to provide new users with the 

software to accomplish their objectives. He knew it would be difficult to 

sell computers without application materials, and without training on how 

to use the systems. As others in the company designed the hardware and 

logic of the EDVAC II, as it was then called, Mauchly focused his 

attention on software questions.

2

 There was no organized software and 

applications department during 1947; Mauchly designed the organization 

as time transpired and machine design determined needs. Nevertheless, in 

early 1947, he began to recruit a staff of mathematicians interested in 

coding for a new department he intended to organize as president of the 

company.  

 

The early members came with experience in coding for various computer 

systems including ENIAC. Frances Elizabeth Snyder (later Holberton) 

joined EMCC in 1947. She had already had an illustrious career in 

computing, as one of the original members of the ENIAC computing 

group. In 1942, shortly after graduation from the University of 

Pennsylvania with a degree in journalism, she joined the Philadelphia 

Computing Unit at the Moore School. This group worked on problems 

associated with the tables being produced by the Aberdeen Proving 

Ground. Snyder, and another member of the group, Betty Jean Jennings 

(later Bartik), developed a trajectory program used to control the operation 

of the ENIAC during the public demonstration in February 1946. In 1947, 

Snyder transferred to Aberdeen when the ENIAC was moved there. She 

served EMCC as a part time consultant in February through April 1947, 

and in July of 1947, she left the civil service and became an employee of 

EMCC.

3

 She stated later that she had actually asked Mauchly if she could 

join EMCC.

4

  

 

Jean Jennings Bartik graduated from Missouri State Teachers College 

(now University) in 1945, with a mathematics major. She studied analytic 

geometry, trigonometry, and physics. In summer 1944, she worked at 

Pratt-Whitney Aircraft in Kansas City on engine work, but did not want to 

engage in that type of work after graduation. She applied for a job at the 

Aberdeen Proving Ground for their location at the University of 

Pennsylvania. After two months, she received a letter offering her the job 

and she left for Philadelphia the next evening. Not long after, Jennings 

applied for a position as an ENIAC coder, and she, Snyder, and several 

others were chosen and sent to Aberdeen in June 1945 for training. After 

several months at Penn coding for the ENIAC, Jennings was selected to 

---

Iterations – Norberg – Software Development at EMCC 

 4 

head a group to generate programs to turn ENIAC into a stored-program 

system. She decided not to move to Aberdeen when ENIAC moved there, 

and after completing this programming work she accepted a job at EMCC 

in early 1948.

5

 

 

With these accomplished mathematicians, the department immediately 

engaged in a number of projects while they acquired knowledge of the 

needs of users to whom they expected to sell UNIVAC systems. Many 

conversations occurred between Mauchly and Snyder, on the one hand, 

and prospective customers on the other. Snyder remembered early trips to 

the Census Bureau, NBS, and Martin Marietta. Later, she visited many 

UNIVAC I sites for consultation on programming problems.

6

 As she 

pointed out in her 1983 interview for the Charles Babbage Institute (CBI), 

she had no training in computing, no courses in formal logic. Indeed, 

when she interviewed potential programmers for EMCC and elsewhere, 

she asked such questions as “Did you like plane geometry?” in an effort to 

establish whether the person had a puzzle-solving ability, which might 

make them good programmers.

7

 But this line of questioning did not 

always produce the desired result. In an interview in 1990, Jean Jennings 

reported that in EMCC 

 

We used to argue about what kind of a person [made the best programmer]. We 

had this little test that Art Katz worked out. We used to give these people this 

little test… I don’t know what anybody else used, but their enthusiasm for doing 

something new was what always impressed me personally. …Hildegard 

Nidecker came along, who had much experience in doing calculations for the 

Army. And she flunked his test, so nobody wanted to hire her. Then he [Katz] 

said, ‘this is ridiculous. This just proves to me that the test is ridiculous. We 

know that she is going to do a good job,’ and in fact [we hired her] and she 

retired from UNIVAC [in the 1980s]. So the truth is we did it by the seat of our 

pants, and I personally did it if I liked the person.

8

  

 

Whatever the evaluation scheme, an effective group was assembled at 

EMCC. 

 

The applications group grew slowly between 1947 and 1950. M. Jacoby 

joined the firm in December 1947. Dr. Arthur Katz came to EMCC shortly 

thereafter in February 1948 and, as mentioned, Jean Jennings became an 

employee at the end of March 1948. Four more people joined in the 

second half of 1948 and early 1949—M. League, V. Hovsepian, Hubert 

M. Livingston, and Arthur J. Gehring. Besides the special studies for 

prospective customers noted above, Snyder worked closely with Mauchly 

on the early codes (instruction sets) for EDVAC II and BINAC, codes C-1 

through C-5. More special studies involved work for the Army Map 

Service, Oak Ridge, and Glenn L. Martin Company, all completed by 

early 1949. Snyder worked on a floating decimal routine (at one point 

along with Katz), a reciprocal routine, double precision operations, and 

reciprocal square roots. Mauchly, while working with Snyder on the code, 

---