Everything below is "as far as Maple is concerned".

The following indicates that infinity does not satisfy the property of being real.

> is(infinity,real);
                                     false

But this next pair indicates that infinity is real. They show that infinity satisfies the nonstrict inequality relation equivalent to the 'real' RealRange, from which it would follow that infinity is in the 'real' RealRange.

> RealRange(-infinity,infinity);                                                            
                                     real

> convert(x::'RealRange'(-infinity,infinity),relation);
                       And(-infinity <= x, x <= infinity)

Another way to look at the discrepancy is to consider whether that RealRange is intended to be open or closed. And this next pair seem to confirm that Maple is confused about it.

> `property/ConvertRelation`(x < infinity);
                                    x, real

> `property/ConvertRelation`(x <= infinity);
                                    x, real

The type system seems to have been taught the difference between numeric and extended_numeric. But the property (and `is` and `assuming) system appears not to have had similar treatment.

Comments?

acer