On 7/2/2012 9:59 AM, Barton Willis wrote:
> Richard Fateman wrote
>> We should agree that
>> integrate(f(t,x), t, a ,b)
>> has the same meaning as
>> integrate(f(gensym1,x), gensym1, a, b).
> If integration calls asksign on the integration variable, the user will be quizzed on the sign of a gensym.
Either a gensym or not, it should not matter.
That is, the integration variable is "bound" and
integrate(f(t),t,a,b) is semantically identical to
integrate(f(gensym1),gensym1,a,b).
Any question about gensym1 is just as illegitimate as the corresponding
question about t.
We know that a<=t<=b.
If a path is part of the definite integration, we have to find a better
way of describing it than
talking about an otherwise inaccessible bound variable.
> And that's not nice. I know of no easy solution to this. Maybe this could be
> fixed with some special alias mechanism for asksign. Actually asksign should never be called on the integration variable of a definite integral--questions, if any, should be about
> integrand parameters or limits of integration.
YES!
>
> --Barton