2.6. Implementation

As the objective of the PhD is to provide an accurate and efficient calculation method based on raytracing and physical optics, the work performed in the course of the thesis reaches beyond the theoretical formulation of the novel approach. The method’s feasibility can only be shown based on its actual implementation and application to practical use cases. In total, eighty-two individual modules comprising more than 46.000 lines of Pascal source code were written for the PhD. This does, of course, comprise much code that is not relevant for the final, functional code of the RadiCal method, such as testing and validation routines as well as obsolete implementations that were replaced with alternative approaches. Hence, regarding the work performed, the finally selected and presented models and code snippets can only be considered the tip of the iceberg. Although applying measures to software development productivity is challenging, an accepted standard publication Main methodological approaches RadiCal, D. Rüdisser 13 (Jones, 2017) states a value of 615 lines of code per month as an industry average for pascal programming, which would total 75 months of programming for the code written during the thesis. The requirements for documentation in the software industry are higher, of course, and hardly comparable; however, the code implemented here was of a comparably complex nature, and apart from the actual coding, many test runs, validation cases and performance optimisation had to be performed. Due to the large extent of the written source code, presenting or discussing the entire implementation is impossible. Nonetheless, code snippets covering the implementation are essential elements of the thesis and are included in most sections. Their purpose is to provide insights regarding the implementation’s structure and highlight essential algorithms or design choices. The validity of the implementation is proven based on carefully selected tests (see above).