[Cado-nfs-discuss] change of semantics of -admin and -incr

Emmanuel Thomé emmanuel.thome at gmail.com
Tue Sep 16 23:10:49 CEST 2014


Sorry for the very late follow-up on this.

By forcing a_d to be a multiple of e.g. 210, we are forcing some good
root properties at inifinity.

Are you saying that the induced benefit is minimal, to the point that
we should not do that ?

E.

On Tue, Sep 9, 2014 at 8:56 AM, Zimmermann Paul
<Paul.Zimmermann at inria.fr> wrote:
>        Hi,
>
> I propose the following change in polyselect2l.
>
> Currently with -admin admin -incr incr, we start with the smallest leading
> coefficient ad for the algebraic polynomial which is >= admin and multiple of
> incr.
>
> I propose to only require ad >= admin.
>
> As a consequence, we could run in parallel for example -admin 1 -incr 32,
> -admin 2 -incr 32, ..., -admin 32 -incr 32.
>
> In our current parameter files, admin is always 0, thus is always multiple of
> incr, and this would have no consequence.
>
> However in cadofactor.py when we cut [admin, admax] in intervals of length
> adrange, we would have to make sure that the start of each interval equals
> admin mod incr.
>
> Opinions?
>
> Paul
> _______________________________________________
> Cado-nfs-discuss mailing list
> Cado-nfs-discuss at lists.gforge.inria.fr
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/cado-nfs-discuss


More information about the Cado-nfs-discuss mailing list