Design of an integrated programming and operating system, Part II: The assembly program and its language
R. B. Talmadge · IBM Systems Journal · 1963
In the past few years, programming methods in general use have tended to emphasize modularity of program design; not only for large, complex programs, but even for those of relatively modest size. A common method for achieving modularity has been to divide the program into (relocatable) segments, because: • Coding and checkout of the individual segments can proceed in parallel, thereby considerably reducing the time between analysis of the problem and availability of a running program. • Modifications to the program, which are usually confined to a few segments, can be accomplished without the necessity of reassembling the entire program. • Duplication of effort can be minimized by taking advantage of libraries of previously checked-out subroutines. Different programming techniques can be used for the various segments, thus allowing the exploitation of the strong points of a particular language or processor.