前もって論文を執筆する技術戦略

ソフトウェアバグ検出への適用

抄録

The value of writing a paper about a research project before doing the research itself, a practice strongly encouraged in the New Manufacturing System Group,has been confirmed in relation to software debugging. In the preliminary stage before the debugging exercise began, putting the process down in writing and structuring and illustrating it in diagrams, charts, and tables clarified its purpose, technical issues, evaluation methodology, and other matters that had been unclear. Although attended by ‘birth pangs', the writing process necessarily called forth creativity to overcome hurdles. As a result, the debugging process was completed without the need to go back and repeat steps, many mistakes were uncovered, and no bugs were reported after the software was released. The greatest achievement, however, lay in demonstrating and confirming the significance of predocumentation, and it also became evident that predocumentation made the entire project more efficient.