Year: 2014
Type: dissertation
Abstract: Documentation is essential to software development. Experienced \nprogrammers know this well from having worked with poorly documented \ncode. They wish to improve their documentation techniques and \nhabits, but there is little consensus for them to follow. Somehow, \nthe many different standards must be compared objectively. This \ndesire motivates my work, which aims to better understand existing \ndocumentation practices. \n \nThis work focuses exclusively on comments within the program \ncode. Programming is a complex human activity, despite a widespread \nmisconception among programmers that writing code is a mechanical \nprocess. This is especially true of comments, where programmers \nexpress themselves freely. My work fills a gap in research on \nsoftware documentation by systematically investigating the comments \nin a unique database of code written by programmers under natural \nconditions. \n \nThe true variety of programming behaviour is surprising. But this \nvariety does not mean that the output of programmers is completely \narbitrary; there are patterns in this data, which my research aims \nto understand. \n \nThis work makes three contributions: \n \nA novel taxonomy of comments developed from the data, which to date \nis the most thorough description of commenting behaviour actually \nexhibited by programmers. \n \nEmpirical hypotheses regarding large scale commenting behaviour, \nwhich were validated on separate test data. These hypotheses describe \nunderlying regularities in programming which appear to transcend \nindividual differences. \n \nThe database of code I collected, which has unique opportunities \nfor further research on software development, and is thus available \nfor use by other researchers.
Source: UWSpace (University of Waterloo)
Author: Simon Benjamin Orion Parent
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2527330806