The publishing criteria for a siggraph paper is that the description has to be detailed enough that it could be implemented by a grad student experienced in the

The publishing criteria for a siggraph paper is that the description has to be detailed enough that it could be implemented by a grad student experienced in the area. It doesn't have to be directly usable. I've heard that working code counts for more that it used to, but the main goal of academic research is the exchange of ideas and techniques, not making working production software.