Re: polrecip for a constant polymomial

[Dirk Laurie <dirk.laurie@gmail.com>]

Ii is not merely inconvenient, it is inconsistent and may even qualify
as being a bug.

Surely the definition of polrecip should be equivalent to

   polrecip(p) = subst(p,x,1/x)*x^poldegree(p)



2018-02-14 0:41 GMT+02:00 Jacques Gélinas <jacquesg00@hotmail.com>:
> It is inconvenient to have to add code to handle constant polynomials.
> Why not return the constant itself ?
>
> # polrecip(2+0*x)
> %1 = 2
> # polrecip(2)
>   ***   at top-level: polrecip(2)
>   ***                          ^-----------
>   *** polrecip: incorrect type in polrecip (t_INT).
>
> Jacques Gélinas, Ottawa
>