This post discusses the optimization techniques in shader programming, particularly focusing on the use of conditional moves and functions like mix() and step(). A user criticizes the argument against optimizing conditional moves, suggesting that a comparison of machine codes for both methods would strengthen the claims. The conversation reflects on how processor and compiler changes impact performance, with historical context on assembly and C optimizations. It raises questions about the ability of compilers to recognize and optimize operations effectively, and whether optimization passes exist in shader compilers to address inefficiencies. Additionally, there is frustration expressed over the persistence of misinformation regarding optimization strategies in shader programming.