Update Singular to 4.1.0p2

[jpflori]

New version at:

• ​http://www.mathematik.uni-kl.de/ftp/pub/Math/Singular/SOURCES/4-1-0/singular-4.1.0p2.tar.gz

This should fix warnings in debug mode: #21624

[jpflori]

On a ppc64, I get a bunch of

        return new_Expression_from_GEx(self._parent, self._gobj.numer())
    RuntimeError: Exponent overflow Error: Bad Argument Value

which look pynac related and singular unrelated. In fact I seem to get them on a pristine 7.6beta4.

[jpflori]

Or its due to giac being broken on ppc64, no idea.

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​a204dfc

Further easy fixes.

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​e8a265b	Further doctest changes.
​400fce1	Fix Singular's make check with upstream patch.

[jpflori]

I'm not sure this is in a review state as my ppc64 spits a lot of giac related errors, but feel free to finish this ticket as I will be away from the internet for a couple of weeks.

[jpflori]

Anyone? @volker: this should simplify your life as the patch bot building in debug mode won't have singular related errors anymore.

[vbraun]

I'll give it a whirl ...

[vbraun]

Doesn't work:

[singular-4.1.0p2] 1 out of 1 hunk ignored -- saving rejects to file kernel/spectrum/test.cc.rej
[singular-4.1.0p2] patching file kernel/test.cc
[singular-4.1.0p2] Reversed (or previously applied) patch detected!  Assume -R? [n] 
[singular-4.1.0p2] Apply anyway? [n] 
[singular-4.1.0p2] Skipping patch.
[singular-4.1.0p2] 1 out of 1 hunk ignored -- saving rejects to file kernel/test.cc.rej
[singular-4.1.0p2] patching file libpolys/tests/rings_test.h
[singular-4.1.0p2] Reversed (or previously applied) patch detected!  Assume -R? [n] 
[singular-4.1.0p2] Apply anyway? [n] 
[singular-4.1.0p2] Skipping patch.
[singular-4.1.0p2] 1 out of 1 hunk ignored -- saving rejects to file libpolys/tests/rings_test.h.rej
[singular-4.1.0p2] Error applying '../patches/make_check.patch'

[fbissey]

It looks like the patch has been included in upstream tarball. So yes the tarball is the tagged git branch plus this patch at least - dubious. I'll push a fix in a few hours.

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​53385ed

Remove patch already included in upstream tarball

[fbissey]

Fixed.

[vbraun]

Seems to have a different normalization:

sage -t --long src/sage/schemes/affine/affine_morphism.py
**********************************************************************
File "src/sage/schemes/affine/affine_morphism.py", line 672, in sage.schemes.affine.affine_morphism.SchemeMorphism_polynomial_affine_space.dynatomic_polynomial
Failed example:
    F.dynatomic_polynomial([1,1])
Expected:
    (2.00000000000000*x^4 + 5.85640646055102*x^2 + 24.0000000000000)/(x^2 + (-2.00000000000000)*x + 3.46410161513775)
Got:
    (0.125000000000000*x^4 + 0.366025403784439*x^2 + 1.50000000000000)/(0.500000000000000*x^2 - x + 1.73205080756888)
**********************************************************************
1 item had failures:
   1 of  39 in sage.schemes.affine.affine_morphism.SchemeMorphism_polynomial_affine_space.dynatomic_polynomial
    [295 tests, 1 failure, 1.78 s]

[jpflori]

Is that the only failure?

[vbraun]

yes

[fbissey]

One email notification in 10 posts. trac has problems sending email again?

[jpflori]

This failing test was added at #22265. @rlmiller: any idea of what might have changed?

[jpflori]

Ok, only a normalization change, roots did not change. It's just I'm not that good at multiplication by 16.

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​be064f6	Merge remote-tracking branch 'trac/develop' into singular410p2
​fed37d7	Fix dynatomic polynomial test.
​ea637e6	Merge remote-tracking branch 'trac/public/singular410p2' into singular410p2

[embray]

Any chance of getting this merged for 7.6?