This paper describes the DJavadoc system, which was the subject of a real industrial evaluation for several months. From the description provided in this paper, the DJavadoc seems to be promising. However, I could not understand from reading the paper whether or not the DJavadoc system is what Sun Microsystems is using for their formal documentation of the Java Development Kit (JDK) toolkit.

Berglund emphasizes the importance of enabling the electronic documentation of software component libraries, highlighting the role of documentation in understanding library components and their role in the overall software development process. The argument for using the industry as a laboratory is very convincing, and I am very supportive of such an initiative. I also agree with the authors argument that the development of electronic reference documentation that provides efficient reading and coding support for programming purposes has become more relevant and will perhaps become increasingly relevant in the global programming era. This paper is interesting, and it fulfills its purpose. It is also of appropriate length, and includes a sufficient number of references.