Hello, > Does anyone know of any definitive work on the tailcall optimization. > I'm looking for work on what can be risky situations in which tailcall > can cause problems or some documents on what to watchout for while > implementing tailcall. It may be langage (being implemented) dependent. "Modern compiler implementation in ML" by Andrew W. Appel covers this subject at basic level. I cannot think of any serious potential problems. The not serious ones could be: 1. turing out-of-stack into infinite loop, 2. causing a very badly written threaded program to go wild. Regards, Wojtek
