An interesting theory, however I rather suspect it is basically because Limbo had a similar concept for pseudo-enums. I allowed (e.g.) a syntax like: M0,

An interesting theory, however I rather suspect it is basically because Limbo had a similar concept for pseudo-enums. I allowed (e.g.) a syntax like: M0, M1, M2, M3, M4: con (1<<iota); That was taken from one of the Limbo papers.