residue.m & reciprocal functionality?

[Ben Abbott]

I've modified the residue.m script to add compatibility with Matlab's  
equivalent, and would like to solicit opinions with regards to how  
residue.m should behave

The old Octave version worked as

        [R, P, K, E] = residue (B , A, [TOL]);

Where

        R = residues
        P = poles
        K = direct polynomial term
        E = multiplicity of the poles
        B = numerator polynomial
        A = denominator polynomial
        TOL = optional tolerance used to determine multiplicity

Matlab's includes a similar syntax and a reciprocal one

        [R, P, K] = residue (B, A);
        [B, A] = residue (R, P, K);

In Matlab's implementation, the multiplicity vector is not directly  
accessible.

To reconcile the old Octave syntax with Matlab's the optional input,  
TOL, was dropped, and the reciprocal operation was added. So at  
present, my modified residue.m supports the two modes of  
functionality below.

        [R, P, K, E] = residue (B , A);
        [B, A] = residue (R, P, K);

ok, here's the question; Should the reciprocal operation also deviate  
from Matlab and include the multiplicity vector as an optional input?

        [R, P, K, E] = residue (B , A);
        [B, A] = residue (R, P, K, [E]);

... any thoughts/opinions?