‘software gap’, there are many other contributory factors, from the lack of management talents to the employment of unqualified programmers and sheer incompetence in software design.«
McClure: (from Projection versus performance in software production)
»It seems almost automatic that software is never produced on time, never meets specification, and always exceeds its estimated cost. This conference is in fact predicated on this alarming situation. However, on closer inspection the situation does not appear quite so alarming, nor unexplainable, nor incorrigible. The situation is quite analogous to that pertaining in any research and development shop in any line of business whatsoever. The ability to estimate time and cost of production comes only with product maturity and stability, with the directly applicable experience of the people involved and with a business-like approach to project control. The problem with software stems specifically from the refusal of industry to re-engineer last year’s model, from the inability of industry to allow personnel to accumulate applicable experience, and from emotional management.
…
One recent situation is worthy of note. The users of the IBM 7090 used a system called the Fortran Monitor System (FMS) quite satisfactorily for a number of years. Although its facilities were limited, it generally performed as it was supposed to. Very recently, the SDS Sigma 7 was delivered to the accompaniment of howls of anguish because it initially came equipped with only a basic operating system substantially superior to the old EMS. The root problem was that the manufacturer had promised far more and could not deliver on his promises. Did this failure lie in the inability of the software people to produce or in the ability of the sales office to over-promise?«
Gill: (from his Position paper)
»Software is as vital as hardware, and in many cases much more complex, but it is much less well understood. It is a new branch of engineering, in which research, development and production are not clearly distinguished, and its vital role is often overlooked. There have been many notable successes, but recent advances in hardware, together with economic pressures to meet urgent demands, have sometimes resulted in this young and immature technology of software being stretched beyond its present limit.«
Kolence: (from On the interaction between software design techniques and software management problems)
»Programming management will continue to deserve its current poor reputation for cost and schedule effectiveness until such time as a more complete understanding of the program design process is achieved. «
Hastings: Some of the problems are caused by users who like to buy ‘futures’ in software systems, and then ignore the problems inherent in this.
Buxton: There are extremely strong economic pressures on manufacturers, both from users and from other manufacturers. Some of these pressures, which are a major contributory cause of our problems, are quite understandable. For example, the rate of increase of air traffic in Europe is such that there is a pressing need for an automated system of control.
7.1.4. POSSIBLE SOLUTIONS
There were several comments on possible partial solutions, and on the difficulty of finding a simple overall solution to the problems of producing large systems.
Opler: Either of the following two courses of action would be preferable to the present method of announcing a system:
1. Do all development without revealing it, and do not announce the product until it is working, and working well.
2. Announce what you are trying to do at the start of the development, specify which areas are particularly uncertain, and promise first delivery for four or five years hence.