interred should reduce non-leading terms, if there is option redSB and a global order

[Simon King]

The following occurs with Singular-3-1-1-4:

> ring r = 2,(c_4_0, a_1_0, a_1_1),M(4,1,1, -1,0,0, 0,-1,0);
> option(redSB);
> option();
//options: redSB redTail redThrough redefine loadLib usage prompt
> interred(ideal(a_1_0^2,a_1_0*a_1_1+a_1_0^2));
_[1]=a_1_0^2
_[2]=a_1_0*a_1_1
> qring q = std(ideal(a_1_1^2+a_1_0*a_1_1+a_1_0^2,a_1_0^3));
> option();
//options: redSB redTail redThrough redefine loadLib usage prompt
> interred(ideal(a_1_0^2,a_1_0*a_1_1+a_1_0^2));
_[1]=a_1_0^2
_[2]=a_1_0*a_1_1+a_1_0^2

According to the manual, interred should reduce the non-leading terms, if option redSB is used and the order is global (quote: "in the case of a global ordering (polynomial ring) and option(redSB);"). Or is the situation different in quotient rings?

Is there a way to force complete interreduction even if interred wouldn't do it by default?

[hannes]

fixed with rev. 14420