Discussion related to cado-nfs

[paul zimmermann <Paul.Zimmermann@inria.fr>]

       Greg,

> Date: Tue, 1 Sep 2015 19:57:25 -0500
> From: Greg Marks <marks@member.ams.org>
> 
> Dear Paul,
> 
> Sorry for my stupid question; I should have paid closer attention to
> the traceback.  The problem occurred because I had neglected to change
> the slaves.scriptpath variable in my parameters file.

there is no stupid question, only stupid answers...

> I am happy to report that, after downloading and compiling the git
> version just now, everything works.  I got the freerel routine to use all
> 8 of my cores.  (Also makefb used all 8 cores--I don't recall whether
> that was the case with version 2.1.1--although that step is quite fast
> in any event.)  As I write this, the git version has reached the lattice
> sieving stage without any problems.

good to hear!

> Sincerely,
> Greg 
> 
> P.S. To anyone reading this message on the cado-nfs-discuss list: it
> is perhaps worth pointing out that the params/params.cXXX files in the
> git version are different from those in cado-nfs-2.1.1, and a custom
> parameters file modeled on one of them should be updated accordingly.
> For example, the tasks.linalg.bwc.mn parameter seems to be obsolete;
> the git version uses tasks.linalg.bwc.m and tasks.linalg.bwc.n.

indeed, parameter files might change from version x.y.z to u.v.w of cado-nfs.
In particular in the future we plan to provide default parameter files only
for digit sizes multiple of 5, for example params.c155 could be used from 153
to 157 digits.

Paul