General purpose programming systems

The most feasible and general method of logarithm computation for binm'y computers seems to be ~o complete logs initially and then convert to either log ~ or log~0 by a single end multiplication in floating pob~t. ~,mee in binary floating point • 5d¢¢~m~l _< M = ,l[xxxlxxxx... < 1.0d~°~m~l, an 8-position table w~ chose~ to represent the 3 enclosed bits above. Address formation may thus be done by extraction or ~ome other direct method. The basic consideration in the determination of a value for ,~ was the eliminatk)n of the 4th order term by Tchebysheff relaxation, still maintaining an error less than 1 in ~.}~e 27th bit for the logo. (This was chosen for the 27 bit mantissa of a floating point number of the 709 anti may be varied for other machines.) A = .0390625d~¢ was chosen for the table below, being the ia~gest 6 his binary fraction less than the 5 which meets the above specifications. In these remarks, I do not propose to describe an automatic coding system. My aim is, rather, to pre-~e:nt aa approach to automatic coding or, if you prefer, to present a motivation which has guided us in the development of the three actual systems for Remington Rand computers. The earliest of these three sy~:em~, developed for the UNIVAC I, has been known by the name Generalized Programming. (This system w~ recently renamed FLEXIMATIC by the Remington Rand Sales Office.) In operation for approximately two years, GP is currently employed at a considerable number of commercial installations. The seco~d system, Generalized Programming Extended (GPX), is a new version of FLEXIMATIC for the UNIVAC II; it has considerably increased powers in comparison with its ancestor.