@incollection{bemer_print_1957,
	address = {Philadelphia, PA},
	series = {Journal of the {Franklin} {Institute} {Monograph}},
	title = {Print 1 — {An} {Automatic} {Coding} {System} for the {IBM} 705},
	number = {3},
	booktitle = {Automatic {Coding}: {Proceedings} of the {Symposium} on {Automatic} {Coding}, {January} 24-25, {Franklin} {Institute}, {Philadelphia}},
	publisher = {The Franklin Institute},
	author = {Bemer, Robert W.},
	month = apr,
	year = {1957},
	pages = {29--38}
}

Coding efficiency and time savings. At Westinghouse Electric, a magnetic field parameter study was coded in a week using 1300 IBM 705 operation code instructions. After a day’s instruction in the PRINT 1 automatic coding system for the IBM 705, the same programmer recoded the program in twenty minutes using only sixty instructions. This suggests a ratio of forty to one between IBM 705 operation code instructions and PRINT 1 autocoding instructions, and a ratio of one hundred hours to one. Although this might be an exceptional case, other reports suggest “that detailed coding (not programming) effort may be reduced by a factor of 10 through the use of this and other modern automatic coding systems such as FORTRAN, the Carnegie Tech Compiler, BACAIC, B-Zero and OMNICODE” (30).

Instructional efficiency. Two employees of National Supply without any computer experience attended the PRINT 1 training class at Westinghouse and were able to write successful programs in PRINT 1 without ever having learned any IBM 705 operation codes (30).

Nomenclature. “Certain definitions are adopted from the 704 FORTRAN system in order to understand the hybrid operation of the PRINT system. A program written in the synthetic automatic coding language is called a ‘source’ program. It is processed by a translator to produce an ‘object’ program, which may be produced in either a machine language form, a still symbolic intermediate language such as that of an assembly program, or pseudo-instructions for minimum interpretation. PRINT falls into the last category” (35).

PRINT itself. “Although interpretive in execution, meaning that the required machine language instructions have certain portions fabricated while the problem is running, PRINT is not equivalent to the usual interpretive program of early days in computers. PRINT language is freely and descriptively symbolic, much the same as any compiler, and instructions do not bear a recognizable resemblance to the object pseudo-instructions produced by the pre-editing, or translating, process. Thus the name — PRe-edited INTerpretive. Pre-editing does both assembly and conversion of all components of the synthetic instruction to a pseudo-instruction in a form most rapidly used at execution time” (35).