turning a sequence of epxressions into a function

I am new to Maxima, and I am trying to correct the subtle things that I am 
doing wrong.

If you have the time, perhaps you could examine some code for me.

I have the start of a program that works, when I try to turn part of the 
code into a function, I fail.

The first snippet works just fine, I am plotting a graph knowing only the 
curvature, an
initial point and direction.  The following is somewhat contrived, but it is 
a start, and works fine.

I start off by defining the curvature k(a), the arclength, and a function 
called the witch of agnesi.
I then create an array called points, big enough to hold all my increments, 
calculate the initial point and direction,
then fill in the rest of the points using the curvature, moving along the 
arc length.

this works, but I want to turn it into a function.


********************************************

kill(all)$

k(a) := diff(a,x,2)/(1+diff(a,x,1)^2)^(3/2)$
arclength(f,a,b) := quad_qags(sqrt(1+diff(f,x,1)^2),x,a,b)[1]$
witch(a) := (8*(a)^3)/(x^2+4*(a)^2)$


c : witch(1)$
a : -15$
b : 15$
increment : 0.01$


length : arclength(c,a,b)/increment$
points:makelist([0,0],x,0,floor(length))$


f : k(c)$

angle : float(atan(diff(c,x)))$
x : a$
y : ''c$
angle : ''angle$
points[1] : [x,y]$

for i : 2 thru length+1 do (
    angle : angle + float(''f)*increment,
    x : x + cos(angle)*increment,
    y : y + sin(angle)*increment,
    points[i] : [x,y]
)$

plot2d([discrete, points]);

********************************************


The problem that I am having is turning this into a function. Here is what I 
believe might be the right thing.

********************************************

kill(all)$

k(a) := diff(a,x,2)/(1+diff(a,x,1)^2)^(3/2)$
arclength(f,a,b) := quad_qags(sqrt(1+diff(f,x,1)^2),x,a,b)[1]$
witch(a) := (8*(a)^3)/(x^2+4*(a)^2)$

get(c,a,b,increment) :=
block( [length,points,f,x,y,angle],
    length : arclength(c,a,b)/increment,
    points:makelist([0,0],x,0,floor(length)),
    f : k(c),
    angle : float(atan(diff(c,x))),
    x : a,
    y : ''c,
    angle : ''angle,
    points[1] : [x,y],
    for i : 2 thru length+1 do (
        angle : angle + float(''f)*increment,
        x : x + cos(angle)*increment,
        y : y + sin(angle)*increment,
        points[i] : [x,y]
    ),
    points
)$

c : witch(1)$
a : -15$
b : 15$
increment : 0.01$

P : get(c,a,b,increment)$

plot2d([discrete, P]);

********************************************


Maxima does not like this at all.   Any help would be appreciated.  I think 
it may have to do with evaluating things wrong.