Using the new 1.2.3, in ginsh:

series((1+x)^y,x,5); 1+Order(x^5)

while

series(exp(y*log(1+x)),x,5); 1+(y)*x+(-1/2*y+1/2*y^2)*x^2+(1/3*y-1/2*y^2+1/6*y^3)*x^3+(1/24*y^4-1/4*y+11/24*y^2-1/4*y^3)*x^4+Order(x^5)

My understanding is the latter is correct as it gives 1/n! the Stirling1 triangle and substituting a number for y gives binomial coefficients. Maybe any expansion of a power can be expressed by such a construction, even in the GiNaC source, so a patch would be straightforward? ralf