Abstracto

On the Design of an Object Oriented Programming Language

Mokua Richard

This paper presents work done to address two issues of interest to both Programming Language (PL) theory and software development: 1) The inadequacies of mainstream Object Oriented Programming Languages used in the software industry such as Java, C# and C++ and 2) The design and implementation of a statically typed Object Oriented Programming Language that addresses some of the issues identified above. Research was conducted through critical analysis of existing Object Oriented Programming Languages (OOPL) as well as a literature review of journal and conference publications in that area. The aim was to elicit evidence of PL constructs that had been found through previous experience to lead to poor Software Engineering practices such as increased amount of bugs, poor maintainability, late (i.e. runtime) detection of errors, poor usability and low programmer productivity. This work has produced key benefits that include a deeper understanding of PLs specifically OOPLs, and an improved comprehension and appreciation of the nuances of PL design. The findings have the potential to benefit PL researchers and designers in various ways. We consider that the contributions of this work are that a list of the language constructs (e.g., Static Variables, Lack of Object Level Encapsulation, Presence of Primitive Types) that seem to lead to poor Software Engineering practices with current OOPL have been identified. A further significant contribution is the production of a new OOPL designed to act as proof of concept to illustrate how these issues can be addressed.