David Johnson wrote: > My opinion is that "deliberately obfuscated source code" should be decoupled > from documentation. The quality and state of documentation is very > subjective, and should not be a part of the OSD.
I have to agree with David. The documentation quality of the source code is orthogonal to the availability of source code, and thus has nothing to do with the OSD. Trying to establish what documentation quality is, is difficult in the first place. Firstly, people differ in intelligence and experience, so what is obfuscated to one person, may be obvious to another. Secondly, should the quality be judge on the choice of human language? For example, if a russian developer releases source code with comments in Russian, can I claim that he is deliberately obfuscating the source code? Can the russian developer claim that all source code with English comments are obfuscated to him? Thirdly, the source code may implement algorithms or domain knowledge that is inherently difficult to understand, and which would require a book-sized explanation. Would it be considered compliant with the OSD to refer to a (commercially available) book? If not, how does the developer avoid infringing the copyright of the book author while adhering to the suggested OSD documentation requirements? I am sure that there are other concerns as well; the above was simply off the top of my head. I understand the good intentions behind the proposal, but I definitely see it as a slippery slope. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3