Inconsistency when using previously defined variables as arguments to functions



maxima-bounces at math.utexas.edu wrote on 01/26/2011 10:08:18 AM:

> The idea of sum etc taking lambda expressions as arguments is 
intriguing.

It is intriguing, but for double sums, sum would need to return a
lambda form--otherwise sum(sum(lambda([k], k),1,n),1, m) wouldn't
work. Alternatively, we'd need a non-nested way to represent multiple
sums.

I'm OK with sum(lambda([k], k),1,n) --> lambda([k],k*(k+1)/2), but
that would break stuff.

---Barton