Stavros Macrakis

noundisp for operators

If they really want it to do nothing, they can write 'makelist(...), which can later be evaluated with ev/nouns, or '(makelist(...)) which can later be evaluated with ev. I wouldn't say that programming constructs "don't honor simp". I'd say that simp should be irrelevant to their functioning. Any mathematical/simplifying steps within them do honor simp. Unfortunately, it's hard to implement that consistently. What do we do with simp:false;makelist(i,i,sin(0),1)? In principle, sin(0) shouldn't simplify...

I'm OK with binding simp to t just for the arithmetic. After all, makelist is a programming construct, not a mathematical construct. If it were a mathematical construct, it would return a noun form when it couldn't carry out the arithmetic for whatever reason: sum(f(x),x,a,b) => 'sum(f(x),x,a,b) makelist(f(x),x,a,b) => ERROR The same argument applies to do, which could also bind simp to t for the loop variable calculations. The current behavior of do is a bit mysterious (I didn't look at the code):...

I'm OK with binding simp to t just for the arithmetic. After all, makelist is a programming construct, not a mathematical construct. If it were a mathematical construct, it would return a noun form when it couldn't carry out the arithmetic for whatever reason: sum(f(x),x,a,b) => 'sum(f(x),x,a,b) makelist(f(x),x,a,b) => ERROR The same argument applies to do, which could also bind simp to t for the loop variable calculations. The current behavior of do is a bit mysterious (I didn't look at the code):...

I'm OK with binding simp to t just for the arithmetic. After all, makelist is a programming construct, not a mathematical construct. If it were a mathematical construct, it would return a noun form when it couldn't carry out the arithmetic for whatever reason: sum(f(x),x,a,b) => 'sum(f(x),x,a,b) makelist(f(x),x,a,b) => ERROR The same argument applies to do, which could also bind simp to t for the loop variable calculations. The current behavior of do is a bit mysterious (I didn't look at the code):...

makelist uses the simplifier to perform arithmetic in order to cover all kinds of number, e.g., makelist(i,i,2/3,5.0b0) => [2/3,5/3,8/3,11/3,14/3] makelist(i,i,-2.1b0,0,7/8) => [-2.1b0, -1.225b0, -3.5b-1] As an incidental bonus, it handles some non-numeric cases: makelist(i,i,n,n+2) => [n, n+1, n+2] makelist(i,i,sqrt(%e),%pi) => [sqrt(%e), sqrt(%e)+1] I say incidental because it doesn't handle the general symbolic case, where the termination condition requires symbolic comparisons. But I don't think...

Hmm, not consistently: integrate(x,x,0,1+zeroa) => result involving zeroa integrate(x,x,1,atan(inf-1)) => integral is divergent integrate(x,x,1,limit(atan(inf-1))) => %pi^2/8-1/2 integrate(exp(-x),x,zeroa,inf) => gamma_incomplete(1, zeroa) integrate(exp(-x),x,limit(zeroa),inf) => 1 Should it?

sign(zeroa) should be pos