[GiNaC-list] quo behaviour

[Ralf Stephan]

Hello,
sorry for bothering you so much, you probably think by now I get 
paid for this. 

The tutorial says that parameters to quo() should be univariate 
polynomials. Is this really necessary, esp. if you also get the
main variable as parameter? 

Even if I accept the univariateness for parameters[*], the behaviour
of the function towards multivariate input is inconsistent, as the
ginsh session shows:

> quo(y,y*x,y);
FAIL
> quo(y*x,y,y);
x

Thanks,
ralf

[*] I do not accept it, as I understand ginac from its interface to be 
inherently multivariate, except for the diff function, and even that is
disputable.