@incollection{katz_systems_1957,
	address = {Philadelphia, PA},
	series = {Journal of the {Franklin} {Institute} {Monograph}},
	title = {Systems of {Debugging} {Automatic} {Coding}},
	number = {3},
	booktitle = {Automatic {Coding}: {Proceedings} of the {Symposium} on {Automatic} {Coding}, {January} 24-25, {Franklin} {Institute}, {Philadelphia}},
	publisher = {The Franklin Institute},
	author = {Katz, Charles},
	month = apr,
	year = {1957},
	pages = {17--27}
}

Growth of “automatic coding.” “In May 1952, the first compiler, A-Zero, was presented to a critical and unreceptive computer world. In five short years, the field of automatic coding has grown so tremendously, that today there are more compilers than there are computers […] Whereas just a few years ago it was unusual to have one paper on automatic coding presented at a computer symposium, today entire symposiums such as this are dedicated to the discussion of new developments in the field of automatic coding and its applications” (17).

Approaching universality. Pseudo-codes used by the new compiler systems “more and more closely approach the languages familiar to the users” (17). Formula translator systems use algebraic expressions, while the “B-Zero business compiler employs English words as its pseudo-code” (17). Compiled code is no longer necessarily less efficient than hand-written machine code (18).

Labor. “The increasing demand for programmers in the face of the current shortage of trained scientific personnel, has dictated that non-specialized personnel be readily able to use the computers as a tool. The new compilers make it possible for mathematicians, physicists, statisticians, business methods and systems personnel, and accountants to use the computer with a minimum of formal training” (17). At the same time, “[t]he people who can best do scientific work are being spread out over many fields — all growing and requiring more personnel, so it seems a logical conclusion that the number of available trainees must decrease” (22). [This last argument appears in the transcript of discussion following the paper itself, in answer to the following question by Millard H. Perstein: “[W]hy do you think the number and caliber of programmers will decrease in the future?” The exchange concluded thus: Perstein: “I see what you mean. It is a sort of relative thing. You mean, there won’t be as many relative to the demand.” Katz: “Yes, the relative number will be smaller” (22).]

Machine independence. “The newer pseudo-codes have become independent of any particular computer. Thus, once a problem is defined in pseudo-code, compilers for any one of a line of computers can be used to produce programs for themselves” (17).

Importance of subroutine generators. “One of the most important contributions made to the field of automatic coding has been the concept of generators” (18), which created subroutines customized according to parameters in the pseudocode.

New difficulties in debugging. A programmer “who has defined his problem in a pseudo-code is confronted with an error in a machine-coded program […] knows little or nothing about the program he must debug” (19). To address this problem, “systems must be devised that will permit all debugging to be done at the pseudo-code level” (19).