An assessment of reuse technology after ten years

Abstract

Abstract

Finding reusable software components in large systems

“…The best validation of the subsystem concept was that on their own volition developers made subsystem diagrams; put them on their office walls; and used them for reference, planning, and training. The architectural concept of subsystems is an important one for forward synthesis systems such as Draco that purport to address the building large systems [10]. We also believe that the subsystems can be a basis for converting an old system to object-oriented programming as discussed in section 4.4.4.…”

Reuse so far: phasing in a revolution

“…The problem with this approach is that it combines the developing system version / configuration space [3] with those of all the parts you use. In the short-term this is a fine approach.…”

“…As software experts we surly know that our systems requirements, development techniques, tools and hardware all change at a dramatic rate. Many of the concepts I advocate [3] (e.g., structural architecture, version space, configuration space, and specialization) are only for flexibility. The resulting system embodies these concepts but they do not change the system's function.…”

“…In addition, our moderator [1] has challenged us to examine whether any reuse technology makes a revolutionary difference. To some degree I have already done this in my keynote address [3] but Biggerstaff brings up some interesting points.…”

“…This is a safe definition since it has been estimated that software productivity only improved about 3% to 8% per year [5] over a twenty year time frame that included the rise of Software Engineering, high-level languages, and on-line computing. I derived a short lifecycle model [3] using Boehm's data [2] for new development of a large project. The lifecycle phases and their percent of the total effort are:…”