From: fateman@cs.berkeley.edu   
      
   On 8/28/2017 9:30 AM, Nasser M. Abbasi wrote:   
   >   
      
   given this problem:   
      
   INT((a + b*x)/((2 - x^2)*(x^2 - 1)^(1/4)), x)   
      
   it seems to me to be easily decomposed as   
      
   a*INT(1/((2 - x^2)*(x^2 - 1)^(1/4)), x)   
      
   +b*INT(x/((2 - x^2)*(x^2 - 1)^(1/4)), x)   
      
   Next, assuming that the integrals of actual interest are   
   not indefinite, but definite between 2 specific constants,   
   e.g.   
   INT((a + b*x)/((2 - x^2)*(x^2 - 1)^(1/4)), {x,1,4})   
      
   then this and similar problems can be done by numerical quadrature.   
      
   Even if you think that the symbolic form is more general,   
   if you are going to evaluate it at some value for x, consider whether   
   it wouldn't be better to evaluate the integral rather than the   
   elliptic functions.   
      
   Of course doing a symbolic integral has a magical   
   quality to it,  but where should system builders spend their time?   
      
   RJF   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)