RESUMEN

If we look at the tools we use to convert source code into executable format, we can get an accurate picture of the abstraction level that programmers face during construction and maintenance (where the largest chunk of software development effort takes place). Collecting data from the building of 7,000 application programs showed that most CPU time is spent compiling C and C++ code using tools with roots in the 1970s and 1980s. To see order-of-magnitude productivity improvements, we must raise our code's level of abstraction by adopting more modern technologies such as domain-specific languages, general purpose declarative languages, and executable UML.