Dario Teixeira wrote: > Thank you, Romain and Arnaud. With that "list ref" example in mind, it > does make sense for the compiler to play it safe and declare foobar2 to > be non-polymorphic. Moreover, this is one of those issues where I I > suspect that compiler elfs must have pondered already how easy/feasible > it would be to extend the compiler to detect sound instances (such as > foobar2) that could be accepted...
They certainly did: http://mlton.org/ValueRestriction has links to the various papers on the subject (the present scheme was not the first solution for SML, as it notes). David -- Caml-list mailing list. Subscription management and archives: https://sympa-roc.inria.fr/wws/info/caml-list Beginner's list: http://groups.yahoo.com/group/ocaml_beginners Bug reports: http://caml.inria.fr/bin/caml-bugs
