Ranging from posting like "math is too hard," to "let's let the requirements emerge," I always troubled by the lack of attention to the underlying - many times the deep underlying - principles of managing software development projects.
Back in March, Guerrilla Project Management had a wonderful post about the Standish Choas report. But more imporantly are the papers referenced in the post. Rarely and I mean rarely do I encounter commerical IT projects that have discussions around these topics.