Michael J. Synge

I was so impressed and pleased by RDNUM, A. Hassitt's General-Purpose Input Routine [Comm. ACM 7, 6 (June 1964), 350-355], that I have transliterated it into FORTRAN II for the IBM 7094. In doing this I stumbled across solutions to a decimal-to-binary conversion problem that has long bugged FORTRAN. Let me hasten to say, before C. H. Weisert [Letter to the Editor, Comm. ACM 7, 5 (May 1964), 314] castigates me for discussing 7090 problems in your columns, that the matter discussed here is relevant to any machine conversion algorithm.