It doesn't use it just because you set the environment variable. You
can simply remove the ntl-prefix if you want.

Bill.

On 4 Aug, 12:02, "Dr. David Kirkby" <[email protected]> wrote:
> On 08/ 4/10 04:03 AM, Fran ois Bissey wrote:
>
>
>
>
>
> >> On 8/3/2010 7:34 PM, Dr. David Kirkby wrote:
> >>> I think I've discovered something which should be a blocker for 4.5.2.
>
> >>>http://trac.sagemath.org/sage_trac/ticket/9681
>
> >>> Please correct me if I'm mistaken. It would not be the first time I've
> >>> been mistaken over dependencies in the spkg/standard/deps file
>
> >> Thanks very much for noticing this!
>
> >> To any experts, not necessarily excluding David:  Does zn_poly as it's
> >> currently used in Sage depend on NTL?
> > Not an expert. But it looks like if ntl is not present at configuration time
> > zn_poly has a fall back. See include/support.h
> > NTL use is very limited, 2 operations. If NTL is found profile/ntl-profile.c
> > is used which calls the following ntl headers:
> > #include<NTL/lzz_pX.h>
> > #include<NTL/ZZ_pX.h>
>
> I appreciate NTL might not be necessary in general, but the configure script
> actually sets the patch to NTL, with:
>
> configure --gmp-prefix="$SAGE_LOCAL" --ntl-prefix="$SAGE_LOCAL" \
>              --prefix="$SAGE_LOCAL" --cflags="$CFLAGS" --ldflags="$LDFLAGS"
>
> I would think if one sets --ntl-prefix to be "$SAGE_LOCAL", then NTL should be
> present.
>
> What I proposed was adding a dependency in spkg/standard/deps file, to ensure
> ntl was built before zn_poly.
>
> Dave

-- 
To post to this group, send an email to [email protected]
To unsubscribe from this group, send an email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org

Reply via email to