		HaMLet	SML/NJ	MosML	MLton	MLKit	Poly/ML	MLWorks	Alice
		1.3.0	110.59	2.01	051202	4.3.0	5.0	2.0	1.4

The following should be accepted and execute without raising an exception:

abstype(4)	-	-	-	+	-	-	-	+
asterisk	+	-	-	+	+	+	+	+
exhaustive	+	-	+	+	+	-	+	+
flexrecord	+	+	+	+	+	+	-	+
fun-case(5)	-	-	-	-	-	-	-	-
fun-infix	+	+	+	+	+	+	?	+
fun-partial	+	+	+	+	-	+	?	+
functor-poly	-	-	-	+	-	-	?	-
generalise	+	+	+	+	+	-	?	+
id		+	-	-	+	+	+	+	+
open		+	-	+	+	+	+	-	-
overloading	+	-	+	+	+	+	-	+
poly-exception	+	-	-	-	-	-	?	-
replication	+	-	+	+	+	-	?	+
scon		+	-	+	+	-	+	-	+
semicolon	+	+	+	+	+	-	+	+
sharing		+	-	+	+	+	+	?	-
typespec	+	+	-	+	+	-	+	+
undetermined	+	-	+	+	-	+	+	+
valrec		+	-	-(3)	-	-(2)	-(2)	-(3)	-
where		+	-	-	+	+	-	-	-
where-and	+	-	-	+	-	-	-	+
withtype	+	-	+	+	+	+	+	+

The following should be rejected or produce a warning:

abstype2(4)	+	+	+	-	+	+	?	-
dec-strdec(6)	+	+	-	+	+	-	?	+
functor-poly2	+	+	+	-(8)	+	+	+	+
nonexhaustive	+	+	+	+	+	+	+	+
redundant(7)	-	-	-	+	-	-	+	-
tyname(6)	+	-	+	+	+	-	?	-
tyvar-shadowing	+	-	+	+	+	-	-	+
undetermined3	+	+	-	+	+	+	?	+
where2		+	-	+	+	+	+	-	+

It is a matter of interpretation whether the following have to be accepted:

flexrecord2	-	-	-	+	-	+	-	-
undetermined2	-	-	+	+	+	-	?	+


Notes:

(1) Internal compiler error.
(2) Internal runtime exception.
(3) Deviating dynamic behaviour, which can be seen as fixing a bug in the
    Definition.
(4) It is practically impossible to be conforming wrt abstype and abstype2
    at the same time.
(5) It is practically infeasible to be conforming wrt fun-case.
(6) Failure (acceptance) can be seen as a reasonable conservative extension.
(7) Conformance is infeasible with separate compilation.
(8) Only gives a generic warning not specific to the example.
