5. PRODUCTION
5.1. INTRODUCTION
Section 3 of this report contains much material on the place of Production in the total task of designing implementing, delivering, maintaining, etc., a software system. During the conference a working party produced a report, Classification of subject matter, which attempted to list and classify the procedures that constitute the production process and the technical components involved in the production task. This report is reproduced in its entirety in Section 9. The conference covered only a small part of the subject matter listed by the working party, so the organization of this section of the conference report does not exactly follow the classification proposed by the working party.
5.1.1. THE PROBLEMS OF SCALE
The rate of growth of the size of software systems is dramatically represented by figure 6 (prepared by McClure). This shows, on a logarithmic scale, the amount of code (compiled instructions) provided as standard programming support (Type 1 programs in IBM terminology) for a variety of computers. The reader is invited to use this chart in order to make his own predictions as to the amount of programming support that will be provided with systems delivered in 1975, and to speculate on the number of systems programmers that might then be required.
Additional data regarding the size of some current systems was given by Nash, in his paper ‘Some problems in the production of large scale software systems’. The data, which concerns the number and size of modules in an early version of OS/360 (released in the first quarter of 1966) is reproduced below.
Number Number
Component of of
Modules Statements
Data Management 195 58.6 K
Scheduler 222 45.0
Supervisor 76 26.0
Utilities 86 53.0
Linkage Editor 24 12.3
Testran 74 20.4
Sys. Gen. 32 4.4
Subtotal 709 219.7
Assembly E 32 43.0
Cobol E 77 50.6
Fortran E 50 28.7
Sort 175 56.5
Subtotal 334 178.8
TOTAL 1,043 398.5 K.
The situation that has been brought about by even the present level of programming support was described by one conference member as follows.
David: (from Some thoughts about the production of large software systems (2))
»Production of large software has become a scare item for management. By reputation it is often an unprofitable morass, costly and unending. This reputation is perhaps deserved. No less a person than T. J. Watson said that OS/360 cost IBM over 50 million dollars a year during its preparation, and at least 5000 man-years’ investment. TSS/360 is said to be in the 1000 man-year category. It has been said, too, that development costs for software equal the development costs for hardware in establishing a new machine line. The commitment to many