#2272 radcan bug

[Aleksas]

"radcan" bug

We calculate the integral of
(%i1) f:1/((x-1)*sqrt(-8*x^2+4*x+3));
(%o1) 1/((x-1)*sqrt(-8*x^2+4*x+3))
(%i2) assume(x<1)$
(%i3) integrate(f,x)$ sol:factor(%);
(%o4) asin((6*x-5)/(2*sqrt(7)*(x-1)))

Test of solution(the first method):
(%i5) diff(sol,x)-f$ factor(%);
(%o6) 0
(%i7) is(%=0);
(%o7) true

Test of solution(the second method):
(%i8) diff(sol,x)-f$ radcan(%);
(%o9) (2*sqrt(-8*x^2+4*x+3))/(8*x^3-12*x^2+x+3)
(%i10) is(%=0);
(%o10) false

I think that "radcan" has bug