Numerical evaluation of sequence looks suspicious

Hi folks --

 Don't you think the following numerical evaluation of a list smells fishy?

First we make a normal sequence:

(%i88) symbolic_list:makelist(x*sin(%pi/x),x,1,15);

                     3 sqrt(3)                         %pi
 %pi         %pi
(%o88) [0, 2, ---------     , 2 sqrt(2), 5 sin(---), 3, 7 sin(---), 8 sin(---),
                         2                                    5
         7             8
        %pi           %pi           %pi            %pi           %pi
        %pi
9 sin(---), 10 sin(---), 11 sin(---), 12 sin(---), 13 sin(---), 14 sin(---),
         9              10              11            12
13             14
          %pi
15 sin(---)]
          15

(I hope you can see the correct fractions above - if not, I apologize
for misformatting):

Then we ask for the numerical evaluation of the previous list:

(%i89) floating_list:ev(symbolic_list,float);

(%o89) [0, 2, 2.598076211353316, 2.82842712474619, 5 sin(0.2 %pi), 3,
7 sin(.1428571428571428 %pi), 8 sin(0.125 %pi), 9 sin(.1111111111111111 %pi),
10 sin(0.1 %pi), 11 sin(.09090909090909091 %pi),
12 sin(.08333333333333333 %pi), 13 sin(.07692307692307693 %pi),
14 sin(.07142857142857142 %pi), 15 sin(.06666666666666667 %pi)]

So, what's with these sines?

TIA,


Henry Lenzi