Re: [Axiom-developer] Re: hyperdoc

[root]

> Do you have any experience with OpenMath?

The point of debate with openmath is as follows... Consider axiom and
maple or mathematica. Ideally openmath would allow you to take an
axiom expression, export it to maple, and get back an answer so that
the computation was transparent.

But in Axiom most of the semantics of an equation is embedded in the
type. Thus, 

 x:POLY(FRAC(INT)):=1

is not the same as

 x:POLY(COMPLEX(INT)):=1

or 

 x:FRAC(POLY(INT)):=1

or any other variation. I don't know how to design a system-independent
way that Maple can take an axiom computation involving 'x' and return
the correctly typed result. In fact, until we get our act together it
is not always clear what the return type of an expression will be even
within axiom. Eventually we will write out the coerce/convert tree and
be able to tell, of course, but that won't help us with openmath.

Tim