Fortran Automatic Coding System - IBM 709 General Information Manual

Fortran Automatic Coding System
The
IBM·
Mathematical FORmula TRANslating System,
709 FORTRAN, is an automatic coding system for the
IBM
709 Data Processing System. More precisely, it is
a 709 program that accepts a source program written
in the Fortran language, closely resembling the ordi-
nary language of mathematics, and produces a ma-
chine language object program ready to be run on a
709.
The 709 Fortran therefore, in effect, transforms the
IBM
709 into a machine with which communication
can be made in a language more concise and more
familiar than the 709 machine language itself. The
result is a substantial reduction in the training re-
quired to program, as well as in the time consumed in
writing programs and eliminating errors from them.
Among the features which characterize the 709
Fortran system are the following:
The 709 Fortran is available in versions for all sizes
of storage. Each version produces programs which
can be used on any size of 709, provided sufficient stor-
age is available for the object program. Object pro-
grams that are too large for the 709 on which they
are to be used must be subdivided by the user.
Object programs produced by Fortran will gener-
ally be as efficient as those written by experienced
programmers.
The Fortran language is intended to provide facili-
ties for expressing any problem of numerical compu-
tation. In particular, problems containing large sets
of formulas and many variables can be dealt with
46 IBM 709·7090
easily, and any variable may have up to three inde·
pendent subscripts.
The language of Fortran may be expanded by the
use of subprograms. These subprograms may be writ-
ten in Fortran language, and may be called by other
Fortran programs, as well as subprograms. The lan-
guage may be expanded by the use of subprograms to
any desired depth.
For problems in which machine words have a logi-
cal rather than a numerical meaning, the language is
less satisfactory, and difficulties may arise in express-
ing such problems. Nevertheless, many logical opera-
tions not directly expressible in the Fortran language
can be carried out by making use of the provisions
for incorporating library routines.
Prewritten routines, to evaluate functions of any
number of arguments, can be made available for in-
corporation into object programs by the use of any
of several different facilities provided for this purpose.
Certain statements in the Fortran language cause
the inclusion in the object program of the necessary
input and output routines. Those which deal with
decimal information include conversion to or from
the internal machine language, and permit consider-
able freedom of format in data input and output.
Arithmetic in an object program will generally be
performed with single-precision floating point num-
bers. These numbers provide about 8 decimal digits
of precision, and may be zero or have magnitudes
between approximately 10-
38
and 10
38 •
Fixed point
arithmetic for integers is also provided.