Issue No. 02 - February (1996 vol. 29)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/2.485847
<p>Can "lean" software compete in the marketplace? Can useful software be elegant? The author laments that a real jewel of a program--written in a consistent style and free of kludges, with simple, organized components--is hard to find. He also wonders why, after more than 30 years of software construction, the industry doesn't produce more jewels. Creators of elegant systems sometimes write articles to tell how they wrote the software and to suggest how others should use their methods. The author points to the T.H.E. system (named for the Technicsche Hogeschool Eindhoven, where the system was built), which has served him as a source of new ideas and insight for 25 years. He also refers to Niklaus Wirth's recent publications (see Computer, February 1995) as musts for every software designer. However, if published papers do indeed contain the secret of success, where are the jewels? Admired systems are often produced under conditions that are rare in industry. In particular, their designers are free from the constraints limiting those who must sell their products. The author discusses why the recipes of the masters haven't led to more elegant commercial software and closes with some advice for those who would like to produce better software.</p>
D. L. Parnas, "Why Software Jewels Are Rare," in Computer, vol. 29, no. , pp. 57-60, 1996.