Sage 9.1.beta9 released

[Volker Braun]

As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html 

I don't see any major issues for 9.1 any more, so we should soon have the first rc. If you have anything that you thing should go in then now is the time to finish it up ;-)

6a4580546f (tag: 9.1.beta9, trac/develop) Updated SageMath version to 9.1.beta9

437bbd8e43 Trac #28671: fix doctests when DOT_SAGE is symbolic link

6f95bfaa3c Trac #29227: Plotting single text3d results in empty scene with Three.js viewer

9591a58d8a Trac #29399: Put Laurent polynomial/series rings into the category of algebras

23bd1c7112 Trac #29397: Add more cygwin system packages

cfd5769395 Trac #29396: Discriminant of multivariate polynomial over RR or RDF gives an error

efb4d84db1 Trac #29394: Drop the python2 patch sys_path_security-issue_16202

1b175bca33 Trac #29357: Add dependencies that should trigger reconfiguring

eecd68a1a3 Trac #29354: tox.ini: Add slackware environments

d4024bd8fc Trac #29304: tox.ini: Add docker-linuxmint

95ced5fb8a Trac #29301: sage-spkg: Add an option -w for warning only if spkg-check fails

b82db0d66b Trac #29295: Add GitHub CI workflow for cygwin

d3772bce89 Trac #29132: Segfault in NTL::ZZ_pContext::restore during exit

eb65fd7999 Trac #29113: Reimplement `sage -i SPKG` for optional/experimental packages as `configure --enable-SPKG && make build`

1f44f2204c Trac #29104: Add to tox.ini some homebrew toxenvs (with a fresh install not in /usr/local), add homebrew packages

7dc258767b Trac #27824: spkg-configure.m4 for python3

be770ceae1 Trac #27247: modular_decomposition should be able to return a tree

f32a397c64 Trac #25358: Test safe directory without python subprocess

322fa860fa Trac #20469: Implement Ariki-Koike algebras

d7db290643 Trac #29082: Do not stage .pc files in src/lib/, clean old generated *.pc files at 'make distclean', fix 'permission denied' errors

d1aa8f5a30 Trac #29383: Remove obsolete scripts from src/bin/

d4609c12f7 Trac #29376: some details in hyperelliptic curves

50bbc41733 Trac #29362: build/bin/sage-pip-install: Delegate to build/bin/sage-python23

2589227e16 Trac #28917: sagemath should not use math names with underscore when generating Latex

d2405f2963 Trac #28052: Incorporating a function for Tautological Implication in the Logic Module

12abc7cd3f Trac #29381: Move testc*.sh from src/bin to build/bin

f98317329b Trac #29380: Update e-antic to 0.1.5

9399c8688b Trac #29375: Clean up some _element_constructor_() methods

89a1805119 Trac #29368: remove files related to SageNB and deprecated function about temporary files

fa7f062dc8 Trac #29361: openblas spkg-configure.m4: Fix the check for lapack/cblas functions

d26d2ae5b3 Trac #29352: more system pkgs info for Debian and Fedora

65914bbfb5 Trac #29342: pari spkg-configure.m4 should check gphelp more carefully

bb0b6ae990 Trac #29323: Set up cross polytope with both Vrep and Hrep (if backend supports it)

084838dc2b Trac #29090: sage-system-python fixup

675d301be6 Trac #23769: src/bin/sage-starts should be moved to build/bin

28957f3f6a Trac #29287: SPKG type: Make "normal/script/pip" orthogonal to "base/standard/optional/experimental"

22ea385799 Trac #29102: fflas_ffpack installation error on new CPUs with old compilers

3f19c4a831 Trac #29170: Fix R installation errors related to gfortran

557e19a2fc Trac #29353: fix parent of q-Catalan numbers

c480201572 Trac #29335: Fix Singular configure so it accepts NTL installed in nonstandard locations

be1e22c441 (tag: 9.1.beta8) Updated SageMath version to 9.1.beta8

[Matthias Köppe]

On Sunday, March 29, 2020 at 8:59:43 AM UTC-4, Volker Braun wrote:

As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html 

I don't see any major issues for 9.1 any more, so we should soon have the first rc. If you have anything that you thing should go in then now is the time to finish it up ;-)

6a4580546f (tag: 9.1.beta9, trac/develop) Updated SageMath version to 9.1.beta9

Let me remark that this is the first beta that tries to use the system python3 in a venv (#27824) instead of building our own copy of python3. Tests of building from scratch or (after "make python3-clean") would be valuable.

I ran tests (make build ptest) at https://github.com/mkoeppe/sage/actions/runs/65613035 . 

Full logs for all builds are available at this link.

Debian (jessie/buster/bullseye/sid), Ubuntu (trusty/xenial/bionic/eoan/focal)m and LinuxMint (19.3) are mostly clean.

Some of the doctest failures are likely known to distribution packagers already. If fixes for these doctests are available already, now it's time to upstream them.

Some doctest failures, for example ubuntu-bionic-standard (https://github.com/mkoeppe/sage/runs/542655815):

sage -t src/sage/interfaces/r.py  # 1 doctest failed

sage -t src/sage/interfaces/tachyon.py  # 1 doctest failed

sage -t src/sage/libs/glpk/error.pyx  # 1 doctest failed

sage -t src/sage/numerical/backends/glpk_backend.pyx  # 1 doctest failed

sage -t src/sage/tests/cmdline.py  # 3 doctests failed

ubuntu-eoan-standard

https://github.com/mkoeppe/sage/runs/542655826

sage -t src/sage/interfaces/giac.py  # 2 doctests failed

sage -t src/sage/interfaces/r.py  # 1 doctest failed

sage -t src/sage/interfaces/tachyon.py  # 1 doctest failed

sage -t src/sage/libs/eclib/interface.py  # 1 doctest failed

sage -t src/sage/libs/glpk/error.pyx  # 1 doctest failed

sage -t src/sage/numerical/backends/glpk_backend.pyx  # 1 doctest failed

sage -t src/sage/tests/cmdline.py  # 3 doctests failed

debian-buster-standard (https://github.com/mkoeppe/sage/runs/542655851)

sage -t src/sage/combinat/tutorial.py  # 2 doctests failed

sage -t src/sage/interfaces/ecm.py  # 3 doctests failed

sage -t src/sage/interfaces/r.py  # 1 doctest failed

sage -t src/sage/interfaces/tachyon.py  # 1 doctest failed

sage -t src/sage/libs/glpk/error.pyx  # 1 doctest failed

sage -t src/sage/libs/libecm.pyx  # 1 doctest failed

sage -t src/sage/numerical/backends/glpk_backend.pyx  # 1 doctest failed

sage -t src/sage/rings/finite_rings/element_pari_ffelt.pyx  # 1 doctest failed

sage -t src/sage/rings/padics/padic_lattice_element.py  # 3 doctests failed

sage -t src/sage/schemes/hyperelliptic_curves/hyperelliptic_finite_field.py  # 2 doctests failed

sage -t src/sage/tests/book_stein_ent.py  # 1 doctest failed

sage -t src/sage/tests/books/computational-mathematics-with-sagemath/combinat_doctest.py  # 2 doctests failed

sage -t src/sage/tests/cmdline.py  # 3 doctests failed

 

debian-bullseye-standard (https://github.com/mkoeppe/sage/runs/542655863)

sage -t src/sage/interfaces/r.py  # 1 doctest failed

sage -t src/sage/interfaces/tachyon.py  # 1 doctest failed

sage -t src/sage/lfunctions/dokchitser.py  # 2 doctests failed

sage -t src/sage/lfunctions/pari.py  # 1 doctest failed

sage -t src/sage/libs/glpk/error.pyx  # 1 doctest failed

sage -t src/sage/interfaces/maxima.py  # Timed out

sage -t src/sage/numerical/backends/glpk_backend.pyx  # 1 doctest failed

sage -t src/sage/rings/number_field/number_field.py  # 8 doctests failed

sage -t src/sage/rings/number_field/number_field_ideal.py  # 2 doctests failed

sage -t src/sage/rings/number_field/unit_group.py  # 1 doctest failed

sage -t src/sage/rings/padics/padic_lattice_element.py  # 3 doctests failed

sage -t src/sage/rings/polynomial/polynomial_quotient_ring.py  # 2 doctests failed

sage -t src/sage/rings/number_field/number_field_element.pyx  # Timed out

sage -t src/sage/schemes/elliptic_curves/ell_number_field.py  # 3 doctests failed

sage -t src/sage/schemes/elliptic_curves/height.py  # 6 doctests failed

sage -t src/sage/schemes/plane_conics/con_number_field.py  # 1 doctest failed

sage -t src/sage/tests/cmdline.py  # 3 doctests failed

Fedora (26-32), Centos (7-8):

Fedora 26: clean

Fedora 27, 28, 29: Critical test failures in sage.combinat, likely because of symmetrica https://trac.sagemath.org/ticket/29405

Fedora 30 and 31: Doctest build error related to eclib 

Fedora 32: Build errors with various packages. Fedora 32 already ships gcc 10!

Centos 7 and 8: clean

Arch Linux: clean

Slackware: Build error with freetype; https://trac.sagemath.org/ticket/29373 needs review

macOS:

homebrew-macos-minimal:

sage -t src/sage/interfaces/gap.py  # 1 doctest failed

sage -t src/sage/rings/padics/padic_lattice_element.py  # 3 doctests failed

sage -t src/sage/sets/recursively_enumerated_set.pyx  # 2 doctests failed

homebrew-macos-standard:

pynac build error - https://trac.sagemath.org/ticket/29404 needs review

homebrew-conda-forge-minimal:

gfan build error - https://trac.sagemath.org/ticket/28984

homebrew-conda-forge-standard:

numpy build error

[Justin C. Walker]

> On Mar 29, 2020, at 05:59 , Volker Braun <vbrau...@gmail.com> wrote:
>
> As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html
>
> I don't see any major issues for 9.1 any more, so we should soon have the first rc. If you have anything that you thing should go in then now is the time to finish it up ;-)

Built from a fresh clone of the develop tree on macOS 10.13.6 (18-core Core i9). All tests (‘ptestlong’) passed!

Builds on other systems to follow.

Justin

--
Justin C. Walker, Curmudgeon at Large
Director
Institute for the Enhancement of the Director's Income
--
In mathematics you don't understand
things. You just get used to them.
--John von Neumann
--

[John H Palmieri]

On OS X with a homebrew installation of Python 3.7.7 I get some failures, mostly to do with pari, including messages like this:

PARI/GP ERROR:
  ***   bug in PARI/GP (Segmentation Fault), please report.

and this:

fatal error: 'gmp.h' file not found
    #include "gmp.h"
             ^~~~~~~
    1 error generated.

and this:

RuntimeError: command 'gcc' failed with exit status 1
ld: library not found for -lmpfr
clang: error: linker command failed with exit code 1 (use -v to see invocation)

"System" packages being used, according to config.log:

boost, boost_cropped, bzip2, cmake, curl, freetype, gcc, gfortran, git, gmp, iconv, isl, libffi, libgd, libpng, mpc, mpfi, mpfr, mpir, ncurses, ninja_build, patch, pcre,pkgconf, python, sqlite, xz, yasm, zeromq, zlib

gsl, openblas, r, and readline are also installed via homebrew, but ./configure doesn't seem to like them.

Failures:

----------------------------------------------------------------------
sage -t --long --warn-long 72.2 src/sage/schemes/elliptic_curves/ell_number_field.py  # 19 doctests failed
sage -t --long --warn-long 72.2 src/sage/schemes/elliptic_curves/ell_rational_field.py  # 31 doctests failed
sage -t --long --warn-long 72.2 src/sage/misc/sagedoc.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/doctest/sources.py  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/schemes/elliptic_curves/height.py  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/symbolic/expression.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/tests/books/computational-mathematics-with-sagemath/integration_doctest.py  # 9 doctests failed
sage -t --long --warn-long 72.2 src/sage/tests/cmdline.py  # 7 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/modform/element.py  # 44 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/number_field/number_field.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/combinat/tableau_tuple.py  # 4 doctests failed
sage -t --long --warn-long 72.2 src/sage/arith/misc.py  # 8 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/modform_hecketriangle/readme.py  # 20 doctests failed
sage -t --long --warn-long 72.2 src/sage/interfaces/maxima_abstract.py  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/abvar/abvar.py  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/structure/element.pyx  # 44 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/abvar/lseries.py  # 11 doctests failed
sage -t --long --warn-long 72.2 src/sage/doctest/forker.py  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/misc/sageinspect.py  # 35 doctests failed
sage -t --long --warn-long 72.2 src/sage/calculus/calculus.py  # 4 doctests failed
sage -t --long --warn-long 72.2 src/sage/quadratic_forms/genera/genus.py  # 4 doctests failed
sage -t --long --warn-long 72.2 src/sage/geometry/polyhedron/combinatorial_polyhedron/conversions.pyx  # 24 doctests failed
sage -t --long --warn-long 72.2 src/sage/graphs/connectivity.pyx  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/misc/cachefunc.pyx  # 54 doctests failed
sage -t --long --warn-long 72.2 src/sage/misc/cython.py  # 19 doctests failed
sage -t --long --warn-long 72.2 src/sage/plot/line.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/modform/constructor.py  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/misc/randstate.pyx  # 24 doctests failed
sage -t --long --warn-long 72.2 src/sage/modules/free_module_element.pyx  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/schemes/elliptic_curves/BSD.py  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/combinat/partition_tuple.py  # 8 doctests failed
sage -t --long --warn-long 72.2 src/doc/en/thematic_tutorials/explicit_methods_in_number_theory/elliptic_curves.rst  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/repl/ipython_extension.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/interfaces/expect.py  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/interfaces/gp.py  # 110 doctests failed
sage -t --long --warn-long 72.2 src/sage/lfunctions/zero_sums.pyx  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/polynomial/polynomial_ring.py  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/parallel/decorate.py  # 7 doctests failed
sage -t --long --warn-long 72.2 src/sage/schemes/elliptic_curves/lseries_ell.py  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/schemes/elliptic_curves/gp_simon.py  # 5 doctests failed
sage -t --long --warn-long 72.2 src/sage/features/__init__.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/matrix/matrix0.pyx  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/modform_hecketriangle/element.py  # 29 doctests failed
sage -t --long --warn-long 72.2 src/sage/groups/perm_gps/permgroup_element.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/structure/parent.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/tests/parigp.py  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/ext/memory_allocator.pxd  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/ext/memory_allocator.pyx  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/arith/long.pxd  # 14 doctests failed
sage -t --long --warn-long 72.2 src/sage/schemes/elliptic_curves/period_lattice.py  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/interfaces/interface.py  # 23 doctests failed
sage -t --long --warn-long 72.2 src/sage/interfaces/tests.py  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/geometry/polyhedron/combinatorial_polyhedron/polyhedron_face_lattice.pyx  # 9 doctests failed
sage -t --long --warn-long 72.2 src/sage/structure/factory.pyx  # 8 doctests failed
sage -t --long --warn-long 72.2 src/sage/cpython/cython_metaclass.pyx  # 4 doctests failed
sage -t --long --warn-long 72.2 src/sage/cpython/getattr.pyx  # 4 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/real_mpfr.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/lfunctions/dokchitser.py  # 71 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/complex_number.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/misc/superseded.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/libs/pynac/pynac.pxd  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/libs/pynac/pynac.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/misc/lazy_attribute.pyx  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/modform/l_series_gross_zagier.py  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/functions/error.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/misc/inherit_comparison.pyx  # 5 doctests failed
sage -t --long --warn-long 72.2 src/sage/matrix/matrix1.pyx  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/integer_fake.pxd  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/misc/session.pyx  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/misc/misc.py  # 16 doctests failed
sage -t --long --warn-long 72.2 src/doc/en/thematic_tutorials/explicit_methods_in_number_theory/nf_introduction.rst  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/geometry/polyhedron/combinatorial_polyhedron/list_of_faces.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/tests/benchmark.py  # 10 doctests failed
sage -t --long --warn-long 72.2 src/sage/docs/instancedoc.pyx  # 4 doctests failed
sage -t --long --warn-long 72.2 src/doc/fr/tutorial/interfaces.rst  # 2 doctests failed
sage -t --long --warn-long 72.2 src/doc/de/tutorial/interfaces.rst  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/symbolic/constants.py  # 4 doctests failed
sage -t --long --warn-long 72.2 src/sage/libs/glpk/error.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/interfaces/quit.py  # 10 doctests failed
sage -t --long --warn-long 72.2 src/sage/repl/load.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/doc/ja/tutorial/interfaces.rst  # 2 doctests failed
sage -t --long --warn-long 72.2 src/doc/ru/tutorial/interfaces.rst  # 2 doctests failed
sage -t --long --warn-long 72.2 src/doc/en/tutorial/interfaces.rst  # 2 doctests failed
sage -t --long --warn-long 72.2 src/doc/pt/tutorial/interfaces.rst  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/cpython/wrapperdescr.pxd  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/padics/padic_ZZ_pX_CR_element.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/lfunctions/pari.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/misc/nested_class.pyx  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/cpython/string.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/structure/parent_old.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/rings/padics/padic_ZZ_pX_CA_element.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/rings/finite_rings/element_pari_ffelt.pyx  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/padics/padic_ZZ_pX_FM_element.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/modular/modform/eis_series.py  # 6 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/complex_double.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/doc/en/thematic_tutorials/explicit_methods_in_number_theory/l_series.rst  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/quadratic_forms/binary_qf.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/rings/real_double.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/misc/sage_eval.py  # 5 doctests failed
sage -t --long --warn-long 72.2 src/sage/lfunctions/lcalc.py  # 3 doctests failed
sage -t --long --warn-long 72.2 src/sage/matrix/tests.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/modular/buzzard.py  # 2 doctests failed
sage -t --long --warn-long 72.2 src/sage/cpython/debug.pyx  # 1 doctest failed
sage -t --long --warn-long 72.2 src/sage/interfaces/mathematica.py  # 1 doctest failed
----------------------------------------------------------------------

For what it's worth, a Python 2 build on the same machine had no doctest failures. It built its own Python 3 rather than using the system one.

[Dima Pasechnik]

Could it be that gcc in Homebrew gets picked to compile C/C++ code?

> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/15dff294-6f9b-4f86-a171-5b8d54fd1c85%40googlegroups.com.

[Matthias Köppe]

Please post config.log and output of "brew list --versions"

[Matthias Köppe]

By the way, current homebrew does not seem to have Python 3.7.x. 

I am not sure how to test with historical snapshots of the homebrew distribution.

[Dima Pasechnik]

On Mon, Mar 30, 2020 at 1:12 PM Matthias Köppe <matthia...@gmail.com> wrote:
>
> By the way, current homebrew does not seem to have Python 3.7.x.

really?
https://github.com/Homebrew/homebrew-core/blob/master/Formula/python.rb
still shows 3.7.7

> I am not sure how to test with historical snapshots of the homebrew distribution.
>
> On Monday, March 30, 2020 at 1:08:44 AM UTC-4, Matthias Köppe wrote:
>>
>> Please post config.log and output of "brew list --versions"
>>
>> On Monday, March 30, 2020 at 12:36:25 AM UTC-4, John H Palmieri wrote:
>>>
>>> On OS X with a homebrew installation of Python 3.7.7 I get some failures, mostly to do with pari, including messages like this:
>>>
>>> PARI/GP ERROR:
>>> *** bug in PARI/GP (Segmentation Fault), please report.
>>>
>>>>
>
>
>
>
>

> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/54268d37-3512-4bfd-8c66-5ea4502570ea%40googlegroups.com.

[Emmanuel Charpentier]

On Debian testing running on core i7 + 16 GB RAM, upgrading 9.1.beta6 to 9.1.beta9 gets 13 permanent errors :

----------------------------------------------------------------------

sage -t --long src/sage/schemes/elliptic_curves/ell_number_field.py  # 5 doctests failed

sage -t --long src/sage/schemes/elliptic_curves/height.py  # 6 doctests failed

sage -t --long src/sage/rings/number_field/number_field.py  # 8 doctests failed

sage -t --long src/sage/schemes/elliptic_curves/gp_simon.py  # 2 doctests failed

sage -t --long src/sage/rings/number_field/number_field_ideal.py  # 2 doctests failed

sage -t --long src/sage/schemes/plane_conics/con_number_field.py  # 2 doctests failed

sage -t --long src/sage/tests/gap_packages.py  # 1 doctest failed

sage -t --long src/sage/lfunctions/dokchitser.py  # 2 doctests failed

sage -t --long src/sage/rings/number_field/number_field_element.pyx  # Timed out

sage -t --long src/sage/numerical/backends/glpk_backend.pyx  # 1 doctest failed

sage -t --long src/sage/rings/polynomial/polynomial_quotient_ring.py  # 2 doctests failed

sage -t --long src/sage/lfunctions/pari.py  # 1 doctest failed

sage -t --long src/sage/rings/number_field/unit_group.py  # 1 doctest failed

sage -t --long src/sage/libs/glpk/error.pyx  # 1 doctest failed

sage -t --long src/sage/interfaces/tachyon.py  # 1 doctest failed

----------------------------------------------------------------------

Re-running these tests individually gives the enclosed chkerrs.txt.

HTH,

[Marc Mezzarobba]

Volker Braun wrote:
> I don't see any major issues for 9.1 any more, so we should soon have
> the first rc. If you have anything that you thing should go in then
> now is the time to finish it up ;-)

I'd really like #28517 to go in, because it fixes some serious issues
reported by a user attempting to do "real" mathematics with Sage and has
been waiting for close to 6 months... But that depends on someone being
available to review it!

--
Marc

[Matthias Köppe]

On Monday, March 30, 2020 at 1:49:28 AM UTC-4, Dima Pasechnik wrote:

On Mon, Mar 30, 2020 at 1:12 PM Matthias Köppe <matthia...@gmail.com> wrote:
>
> By the way, current homebrew does not seem to have Python 3.7.x.
really?
https://github.com/Homebrew/homebrew-core/blob/master/Formula/python.rb
still shows 3.7.7

You are right. 

[Thierry]

Hi,

On Sun, Mar 29, 2020 at 12:50:18PM -0700, Matthias Köppe wrote:
> On Sunday, March 29, 2020 at 8:59:43 AM UTC-4, Volker Braun wrote:
> >
> > As always, you can get the latest beta version from the "develop" git
> > branch. Alternatively, the self-contained source tarball is at
> > http://www.sagemath.org/download-latest.html
> >
> > I don't see any major issues for 9.1 any more, so we should soon have the
> > first rc. If you have anything that you thing should go in then now is the
> > time to finish it up ;-)
> >
> > 6a4580546f (tag: 9.1.beta9, trac/develop) Updated SageMath version to
> > 9.1.beta9
> >
>
>
> Let me remark that this is the first beta that tries to use the system
> python3 in a venv (#27824) instead of building our own copy of python3.
> Tests of building from scratch or (after "make python3-clean") would be
> valuable.

Do you suggest all of us to make distclean and ptestlong ?

Ciao,
Thierry

>
> I ran tests (make build ptest) at
> https://github.com/mkoeppe/sage/actions/runs/65613035 .
> Full logs for all builds are available at this link.
>

> *Debian* (jessie/buster/bullseye/sid), *Ubuntu*
> (trusty/xenial/bionic/eoan/focal)m and *LinuxMint* (19.3) are mostly clean.

> *Fedora* (26-32), *Centos* (7-8):

>
> Fedora 26: clean
>
> Fedora 27, 28, 29: Critical test failures in sage.combinat, likely because
> of symmetrica https://trac.sagemath.org/ticket/29405
>
> Fedora 30 and 31: Doctest build error related to eclib
>
> Fedora 32: Build errors with various packages. Fedora 32 already ships gcc
> 10!
>
> Centos 7 and 8: clean
>
>

> *Arch Linux*: clean
>
> *Slackware: *Build error with freetype;
> https://trac.sagemath.org/ticket/29373 needs review
>
> *macOS:*

>
> homebrew-macos-minimal:
> sage -t src/sage/interfaces/gap.py # 1 doctest failed
> sage -t src/sage/rings/padics/padic_lattice_element.py # 3 doctests failed
> sage -t src/sage/sets/recursively_enumerated_set.pyx # 2 doctests failed
>
> homebrew-macos-standard:
> pynac build error - https://trac.sagemath.org/ticket/29404 needs review
>
> homebrew-conda-forge-minimal:
> gfan build error - https://trac.sagemath.org/ticket/28984
>
> homebrew-conda-forge-standard:
> numpy build error
>
>
>
>
>
>
>
>
>
>
>
>

> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/63f61392-1ddc-4047-ab52-4a7c8095a58b%40googlegroups.com.

[Dima Pasechnik]

On Mon, 30 Mar 2020 at 21:47, Thierry <sage-goo...@lma.metelu.net> wrote:

Hi,

On Sun, Mar 29, 2020 at 12:50:18PM -0700, Matthias Köppe wrote:
> On Sunday, March 29, 2020 at 8:59:43 AM UTC-4, Volker Braun wrote:
> >
> > As always, you can get the latest beta version from the "develop" git
> > branch. Alternatively, the self-contained source tarball is at
> > http://www.sagemath.org/download-latest.html
> >
> > I don't see any major issues for 9.1 any more, so we should soon have the
> > first rc. If you have anything that you thing should go in then now is the
> > time to finish it up ;-)
> >
> > 6a4580546f (tag: 9.1.beta9, trac/develop) Updated SageMath version to
> > 9.1.beta9
> >
>
>
> Let me remark that this is the first beta that tries to use the system
> python3 in a venv (#27824) instead of building our own copy of python3.
> Tests of building from scratch or (after "make python3-clean") would be
> valuable.

Do you suggest all of us to make distclean and ptestlong ?

in principle distclean should not be needed, but certainly make python3-clean

and installing the corresponding “system” packages.

e.g. I don’t think we tested python (should  be python 3.7, i.e. not the latest) from python.org on MacOS.

Dima

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/20200330134729.wugmv2xoo6leseak%40metelu.net.

[Thierry]

On Mon, Mar 30, 2020 at 10:04:46PM +0800, Dima Pasechnik wrote:
> On Mon, 30 Mar 2020 at 21:47, Thierry <sage-goo...@lma.metelu.net>
> wrote:
>
> > Hi,
> >
> > On Sun, Mar 29, 2020 at 12:50:18PM -0700, Matthias Köppe wrote:
> > > On Sunday, March 29, 2020 at 8:59:43 AM UTC-4, Volker Braun wrote:
> > > >
> > > > As always, you can get the latest beta version from the "develop" git
> > > > branch. Alternatively, the self-contained source tarball is at
> > > > http://www.sagemath.org/download-latest.html
> > > >
> > > > I don't see any major issues for 9.1 any more, so we should soon have
> > the
> > > > first rc. If you have anything that you thing should go in then now is
> > the
> > > > time to finish it up ;-)
> > > >
> > > > 6a4580546f (tag: 9.1.beta9, trac/develop) Updated SageMath version to
> > > > 9.1.beta9
> > > >
> > >
> > >
> > > Let me remark that this is the first beta that tries to use the system
> > > python3 in a venv (#27824) instead of building our own copy of python3.
> > > Tests of building from scratch or (after "make python3-clean") would be
> > > valuable.
> >
> > Do you suggest all of us to make distclean and ptestlong ?
>
> in principle distclean should not be needed, but certainly make
> python3-clean
> and installing the corresponding “system” packages.

I am on the way to test it, however it seems that numpy, scipy etc are
not taken into account.

Ciao,
Thierry

> > https://groups.google.com/d/msgid/sage-release/20200330134729.wugmv2xoo6leseak%40metelu.net
> > .

> >
>
> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/CAAWYfq15dTZk6ohHzVss_UuiU5JFeatD3OE9E78Ee6DERVowSw%40mail.gmail.com.

[Dima Pasechnik]

at the moment no packages of the system python are used.
cf https://trac.sagemath.org/ticket/29023

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/20200330141418.sfd5ruelmlwqmavd%40metelu.net.

[John H Palmieri]

"brew list --versions":

autoconf 2.69
automake 1.16.1_1
boost 1.72.0
cabextract 1.9.1
cimg 2.8.4
cmake 3.16.5
emacs-mac emacs-26.3-z-mac-7.8
fontconfig 2.13.1
freetype 2.10.1
gcc 9.2.0_1 9.3.0
gd 2.2.5
gdbm 1.18.1
gettext 0.20.1
glib 2.64.1
gmp 6.2.0 6.1.2_2
gnutls 3.6.12
gpatch 2.7.6
gsl 2.6
icu4c 64.2
ilmbase 2.4.1
isl 0.22.1 0.21 0.18
jpeg 9d
libde265 1.0.5
libdvdcss 1.4.2
libevent 2.1.11_1
libffi 3.2.1
libidn2 2.3.0
libmpc 1.1.0
libomp 9.0.1
libpng 1.6.37
libtasn1 4.14 4.16.0
libtiff 4.1.0
libtool 2.4.6_1
libunistring 0.9.10
mpfi 1.5.3
mpfr 4.0.2
mpir 3.0.0
nettle 3.4.1
ninja 1.10.0
openexr 2.4.1
openssl@1.1 1.1.1d
p11-kit 0.23.20 0.23.18.1
p7zip 16.02_2
pari 2.11.3
pcre 8.43 8.44
pkg-config 0.29.2
python 3.7.7
python@3.8 3.8.2
readline 8.0.4
shared-mime-info 1.15
sqlite 3.31.1
texinfo 6.7
tox 3.14.5_1
unbound 1.9.4 1.10.0
unrar 5.9.1
webp 1.1.0
x265 3.3
xz 5.2.5
yasm 1.3.0_2

config.log attached.

[John H Palmieri]

I should also mention that I have built every beta from scratch for Python 2 and 3 for quite a while, and these failures are new. So if the gcc in homebrew is getting picked up, it's because of some new change in Sage.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-r...@googlegroups.com.

[Dima Pasechnik]

I vaguely recall seeing similar errors on Homebrew, they were due to
stale Homebrew. Could you try updating it?

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/bca9ad32-ca65-4ca5-a159-e1f97627d087%40googlegroups.com.

[Matthias Köppe]

On Monday, March 30, 2020 at 12:23:00 PM UTC-4, John H Palmieri wrote:

I should also mention that I have built every beta from scratch for Python 2 and 3 for quite a while, and these failures are new. So if the gcc in homebrew is getting picked up, it's because of some new change in Sage.

There was no change regarding use of gcc, and from the log it's clear the XCode's gcc (clang) is used.

[Vincent Delecroix]

Le 29/03/2020 à 21:50, Matthias Köppe a écrit :
>
> Let me remark that this is the first beta that tries to use the system
> python3 in a venv (#27824) instead of building our own copy of python3.
> Tests of building from scratch or (after "make python3-clean") would be
> valuable.

Very disappointing

configure:25182: checking ... whether /usr/bin/python3 is good
configure:25329: result: no, Python 3.8.2 is too recent

Vincent

[Dima Pasechnik]

this has been open for a while:

https://trac.sagemath.org/ticket/27754

Vincent

--
You received this message because you are subscribed to the Google Groups "sage-release" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/68fb2185-903d-1369-f688-3e3e066ff756%40gmail.com.

[John H Palmieri]

I ran "brew upgrade" and it only upgraded a few packages: tox, cmake, boost, automake, and some port of emacs. I rebuilt Sage and it didn't help. (Just to be sure, I also built 9.1.beta8, and that one passes tests.)

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/bca9ad32-ca65-4ca5-a159-e1f97627d087%40googlegroups.com.

[Matthias Köppe]

From your config.log ... rather suspicious:

configure:10189: g++ -std=gnu++11 -o conftest -g -O2   conftest.cpp -lgmp  -lm  >&5
configure:10189: $? = 0
configure:10206: result: -lgmp
configure:10231: will use system package and not install SPKG mpir
configure:10265: checking absolute name of <gmp.h>
configure:10272: checking for gmp.h
configure:10272: result: yes
clang: error: unable to execute command: Broken pipe: 13
clang: error: clang frontend command failed due to signal (use -v to see invocation)
Apple clang version 11.0.3 (clang-1103.0.32.29)
Target: x86_64-apple-darwin19.3.0
Thread model: posix
InstalledDir: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin
clang: note: diagnostic msg: PLEASE submit a bug report to http://developer.apple.com/bugreporter/ and include the crash backtrace, preprocessed source, and associated run script.
clang: note: diagnostic msg: 
********************

PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:
Preprocessed source(s) and associated run script(s) are located at:
clang: note: diagnostic msg: /var/folders/cp/n8wtqs490tq5psknff1hv9qr0000gn/T/conftest-d428b4.cpp
clang: note: diagnostic msg: /var/folders/cp/n8wtqs490tq5psknff1hv9qr0000gn/T/conftest-d428b4.sh
clang: note: diagnostic msg: Crash backtrace is located in
clang: note: diagnostic msg: /Users/palmieri/Library/Logs/DiagnosticReports/clang_<YYYY-MM-DD-HHMMSS>_<hostname>.crash
clang: note: diagnostic msg: (choose the .crash file that corresponds to your crash)
clang: note: diagnostic msg: 

********************
configure:10296: result: ///usr/local/include/gmp.h
configure:10332: result: using GMP-compatible library from the system
## ---------------------------------------------------- ##
## Checking whether SageMath should install SPKG gmp... ##
## ---------------------------------------------------- ##
configure:10415: will use system package and not install SPKG gmp

This seems to come from AX_ABSOLUTE_HEADER 

On Monday, March 30, 2020 at 12:18:41 PM UTC-4, John H Palmieri wrote:

config.log attached.

[Steven Trogdon]

Gentoo

Incremental build 9.1.beta8 -> 9.1.beta9

After make python3-clean && make I have the following failures:

----------------------------------------------------------------------

sage -t --long --warn-long 115.0 src/sage/doctest/test.py  # 1 doctest failed

sage -t --long --warn-long 115.0 src/sage/rings/integer.pyx  # 2 doctests failed

----------------------------------------------------------------------

System python is 3.7.6

The failures:

sage -t --long --warn-long 115.0 src/sage/doctest/test.py

**********************************************************************

File "src/sage/doctest/test.py", line 521, in sage.doctest.test

Failed example:

    if system() == "Linux":

        P = subprocess.Popen(["sage", "-t", "--warn-long", "0", "--memlimit=2000", "memlimit.rst"], stdout=subprocess.PIPE, **kwds)

        out, err = P.communicate()

        ok = ("MemoryError: failed to allocate" in bytes_to_str(out))

Expected nothing

Got:

    Process DocTestWorker-1:

    Traceback (most recent call last):

      File "/usr/lib/python3.7/multiprocessing/process.py", line 297, in _bootstrap

        self.run()

      File "/mnt/backup/storage/sage-git_develop/sage/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 2172, in run

        task(self.options, self.outtmpfile, msgpipe, self.result_queue)

      File "/mnt/backup/storage/sage-git_develop/sage/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 2525, in __call__

        result_queue.put(result, False)

      File "/usr/lib/python3.7/multiprocessing/queues.py", line 87, in put

        self._start_thread()

      File "/usr/lib/python3.7/multiprocessing/queues.py", line 170, in _start_thread

        self._thread.start()

      File "/usr/lib/python3.7/threading.py", line 852, in start

        _start_new_thread(self._bootstrap, ())

    RuntimeError: can't start new thread

**********************************************************************

sage -t --long --warn-long 115.0 src/sage/rings/integer.pyx

**********************************************************************

File "src/sage/rings/integer.pyx", line 3118, in sage.rings.integer.Integer.divisors

Failed example:

    for i in range(20):  # long time

        try:

            alarm(RDF.random_element(1e-3, 0.5))

            _ = n.divisors()

            cancel_alarm()  # we never get here

        except AlarmInterrupt:

            pass

Exception raised:

    Traceback (most recent call last):

      File "/mnt/backup/storage/sage-git_develop/sage/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run

        self.compile_and_execute(example, compiler, test.globs)

      File "/mnt/backup/storage/sage-git_develop/sage/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1123, in compile_and_execute

        exec(compiled, globs)

      File "<doctest sage.rings.integer.Integer.divisors[20]>", line 4, in <module>

        _ = n.divisors()

      File "sage/rings/integer.pyx", line 3188, in sage.rings.integer.Integer.divisors (build/cythonized/sage/rings/integer.c:20544)

        ptr = <unsigned long*>check_allocarray(divisor_count, 3 * sizeof(unsigned long))

      File "memory.pxd", line 87, in cysignals.memory.check_allocarray (build/cythonized/sage/rings/integer.c:46769)

    MemoryError: failed to allocate 33554432 * 24 bytes

**********************************************************************

File "src/sage/rings/integer.pyx", line 3790, in sage.rings.integer.Integer.trial_division

Failed example:

    n = 2 * next_prime(10^40); n.trial_division()

Exception raised:

    Traceback (most recent call last):

      File "/mnt/backup/storage/sage-git_develop/sage/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run

        self.compile_and_execute(example, compiler, test.globs)

      File "/mnt/backup/storage/sage-git_develop/sage/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1123, in compile_and_execute

        exec(compiled, globs)

      File "<doctest sage.rings.integer.Integer.trial_division[10]>", line 1, in <module>

        n = Integer(2) * next_prime(Integer(10)**Integer(40)); n.trial_division()

      File "/mnt/backup/storage/sage-git_develop/sage/local/lib/python3.7/site-packages/sage/arith/misc.py", line 1193, in next_prime

        return ZZ(n).next_prime(proof)

      File "sage/rings/integer.pyx", line 5774, in sage.rings.integer.Integer.next_prime (build/cythonized/sage/rings/integer.c:35865)

        while not p._pseudoprime_is_prime(proof):

      File "sage/rings/integer.pyx", line 5309, in sage.rings.integer.Integer._pseudoprime_is_prime (build/cythonized/sage/rings/integer.c:33383)

        return self.__pari__().isprime()

      File "cypari2/gen.pyx", line 2112, in cypari2.gen.Gen.isprime

    cysignals.signals.AlarmInterrupt

**********************************************************************

2 items had failures:

   1 of  23 in sage.rings.integer.Integer.divisors

   1 of  19 in sage.rings.integer.Integer.trial_division

[Simon King]

Hi!

On 2020-03-30, Dima Pasechnik <dim...@gmail.com> wrote:
> On Tue, 31 Mar 2020, 03:03 Vincent Delecroix, <20100.d...@gmail.com>
> wrote:
>> Le 29/03/2020 à 21:50, Matthias Köppe a écrit :
>> >
>> > Let me remark that this is the first beta that tries to use the system
>> > python3 in a venv (#27824) instead of building our own copy of python3.
>> > Tests of building from scratch or (after "make python3-clean") would be
>> > valuable.
>>
>> Very disappointing

Let me add further disappointments:

After making sure that python3.7.7 is installed on my laptop and after
installing the recommended system packages mentioned in
http://doc.sagemath.org/html/en/installation/source.html#linux-recommended-installation
(but note that libflint-arb-dev isn't available for me in ubuntu, no
idea why), I get this in my config.log:

## Checking whether SageMath should install SPKG python3... ##
## -------------------------------------------------------- ##
configure:25026: checking whether any of sqlite libpng bzip2 xz libffi is installed or will be installed as SPKG
configure:25030: result: yes; install python3 as well
configure:25273: no suitable system package found for SPKG python3

Concerning sqlite, I see this:
## Checking whether SageMath should install SPKG sqlite... ##
## ------------------------------------------------------- ##
configure:24630: checking libsqlite3 >= sqlite3_min_version
configure:24664: g++ -std=gnu++11 -o conftest -g -O2 conftest.cpp -lm4rie -lgf2x -lreadline -L/usr/lib/x86_64-linux-gnu -lcurl -lbz2 -lgmp -lm -lntl -lsqlite3 >&5
conftest.cpp:69:62: fatal error: sqlite3.h: No such file or directory

So, apparently libsqlite3-dev needs to be installed (it isn't mentioned in the
list of recommended packages), the other stuff (libpng, bzip2, xz,
libffi apparently is installed on my system).

After installing libsqlite3-dev, I get another problem:
## -------------------------------------------------------- ##
## Checking whether SageMath should install SPKG python3... ##
## -------------------------------------------------------- ##
configure:25026: checking whether any of sqlite libpng bzip2 xz libffi is installed or will be installed as SPKG
configure:25035: result: no
configure:25045: checking for python3 >= 3.7.3, < 3.8 with modules sqlite3, ctypes, math, hashlib, crypt, readline, socket, zlib, distutils.core
configure:25051: result:
configure:25066: checking ... whether /usr/bin/python3.7 is good
configure:25200: result: no, the version is in the supported range, and the modules can be imported, but distutils cannot build an extension
configure:25066: checking ... whether /usr/bin/python3 is good
configure:25200: result: no, the version is in the supported range, and the modules can be imported, but distutils cannot build an extension
configure:25247: result:
configure:25273: no suitable system package found for SPKG python3

Since it mentions "extensions", I was trying to install cython3. But
this didn't help, distutils still cannot build an extension. So, I tried
to install python3-distutils (but this is not available in ubuntu) and
eventually python3-distutils-extra.

Result:
Still doesn't work.

So, it seems to be that the following updates to the Sage installation
guide should be done:

- Something should be done about libflint-arb-dev
- libsqlite3-dev should be added
- It should be explained what to do concerning the system python3's
distutils' inability to build extensions.

Best regards,
Simon

[Matthias Köppe]

Hi Simon,

On Monday, March 30, 2020 at 6:20:24 PM UTC-4, Simon King wrote:

Let me add further disappointments:

After making sure that python3.7.7 is installed on my laptop and after
installing the recommended system packages mentioned in
http://doc.sagemath.org/html/en/installation/source.html#linux-recommended-installation

This is outdated. We now generate an up-to-date version of this manual section at boostrapping time. Use "cat src/doc/en/installation/debian*.txt" to print it.

Or re-run configure and take a look at the system package installation hints that it prints at the very end.

(but note that libflint-arb-dev isn't available for me in ubuntu, no
idea why),

Likely your ubuntu is too old.

 

> configure:25066: checking ... whether /usr/bin/python3.7 is good 

> configure:25200: result: no, the version is in the supported range, and the modules can be imported, but distutils cannot build an extension 

Post config.log, and we can take a closer look.

[Matthias Köppe]

On Monday, March 30, 2020 at 5:54:19 PM UTC-4, Steven Trogdon wrote:

Gentoo [...]

Please help with https://trac.sagemath.org/ticket/29105 so that builds on top of Gentoo can be properly tested.

[Simon King]

Hi Matthias,

On 2020-03-30, Matthias Köppe <matthia...@gmail.com> wrote:
> This is outdated. We now generate an up-to-date version of this manual
> section at boostrapping time. Use "cat src/doc/en/installation/debian*.txt"
> to print it.
>
> Or re-run configure and take a look at the system package installation
> hints that it prints at the very end.
>
> (but note that libflint-arb-dev isn't available for me in ubuntu, no
>> idea why),
>
>
> Likely your ubuntu is too old.

Thanks. Trying what src/doc/en/installation/debian*.txt tells me:
Note, selecting 'zlib1g-dev' instead of 'libz-dev'
E: Unable to locate package libflint-arb-dev
E: Unable to locate package libplanarity-dev
E: Unable to locate package python3-distutils

Concerning too old:
$ uname -a
Linux klap 4.15.0-91-generic #92~16.04.1-Ubuntu SMP Fri Feb 28 14:57:22 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
$ cat /etc/issue
Ubuntu 16.04.6 LTS \n \l

I posted config.log here: https://users.fmi.uni-jena.de/~king/config.log

Best regards,
Simon

[John H Palmieri]

I see the same in the log file for the Python 2 build. I wonder if this has to do with upgrading Xcode, which I seem to have done 6 days ago. To be clear, yesterday and today I have built  9.1.beta8 + Python 3 and 9.1.beta9 + Python 2, both successfully, despite this crash. It's 9.1.beta9 + Python 3 which is causing problems. It also seems to be using the Apple Python 3, /usr/bin/python3, not the one installed by homebrew.

For what it's worth:

- I don't see this crash message from config.log files from a few weeks ago.

- the message refers to a file

/Users/palmieri/Library/Logs/DiagnosticReports/clang_<YYYY-MM-DD-HHMMSS>_<hostname>.crash,

but there is no such file.

[Matthias Köppe]

Hi Simon,

On Monday, March 30, 2020 at 6:45:17 PM UTC-4, Simon King wrote:

Trying what src/doc/en/installation/debian*.txt tells me:
Note, selecting 'zlib1g-dev' instead of 'libz-dev'
E: Unable to locate package libflint-arb-dev
E: Unable to locate package libplanarity-dev
E: Unable to locate package python3-distutils

That looks as expected for Ubuntu 16.04 (xenial).

 

Concerning too old:
$ uname -a
Linux klap 4.15.0-91-generic #92~16.04.1-Ubuntu SMP Fri Feb 28 14:57:22 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
$ cat /etc/issue
Ubuntu 16.04.6 LTS \n \l

I posted config.log here: https://users.fmi.uni-jena.de/~king/config.log

Are you sure your /usr/bin/python comes from that distribution? As far as I can see, xenial only has 3.5.1-3.

 

[Matthias Köppe]

I meant /usr/bin/python3.7, not /usr/bin/python.

[Justin C. Walker]

> On Mar 29, 2020, at 05:59 , Volker Braun <vbrau...@gmail.com> wrote:
>
> As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html

Built from a fresh clone of 9.1.beta9, on two macOS systems: 10.14.6 (2017 MBP, Quad-core Core i7), and 10.15.3 (2019 MBP, 9-core Core i9).

No problems with the builds, and all tests (‘ptestlong’) passed!

--
Justin C. Walker, Curmudgeon-At-Large
Institute for the Absorption of Federal Funds
--------
If you're not confused,
You're not paying attention
--------

[Matthias Köppe]

On Monday, March 30, 2020 at 6:59:30 PM UTC-4, John H Palmieri wrote:

I see the same in the log file for the Python 2 build. I wonder if this has to do with upgrading Xcode, which I seem to have done 6 days ago. To be clear, yesterday and today I have built  9.1.beta8 + Python 3 and 9.1.beta9 + Python 2, both successfully, despite this crash. It's 9.1.beta9 + Python 3 which is causing problems. It also seems to be using the Apple Python 3, /usr/bin/python3, not the one installed by homebrew.

I'll try upgrading XCode to try to reproduce these errors. (My gcc reports "Apple clang version 11.0.0 (clang-1100.0.33.17) Target: x86_64-apple-darwin19.3.0").

FWIW, I just ran "tox -e local-macos-homebrew-standard -- build ptest" 

(as of https://trac.sagemath.org/ticket/29417 it uses the python3 from homebrew)

and it completed with only 1 doctest error in src/sage/interfaces/gap.py.

Did you follow homebrew's instructions regarding the "keg-only packages"? For example, readline is only installed in /usr/local/opt/readline, not in /usr/local.

Sage now has a helper shell script that you can source to set up the environment for those. 

Try ". .homebrew-build-env " before running configure.

[Simon King]

On 2020-03-30, Matthias Köppe <matthia...@gmail.com> wrote:

I installed python3 after "sudo add-apt-repository ppa:deadsnakes/ppa".
Has that been a mistake? How to undo it?

Best regards,
Simon

[Matthias Köppe]

On Monday, March 30, 2020 at 8:30:34 PM UTC-4, Simon King wrote:

> Are you sure your /usr/bin/python comes from that distribution? As far as I
> can see, xenial only has 3.5.1-3.

I installed python3 after "sudo add-apt-repository ppa:deadsnakes/ppa".
Has that been a mistake? 

I don't know, but I found these instructions: https://launchpad.net/~deadsnakes/+archive/ubuntu/ppa

that also mention what packages should be installed

 

[Simon King]

Dear Matthias,

Excellent! After installing them and doing ./configure, I see this line:
python3-3.7.3.p1: using system package; SPKG will not be installed

So, now I can try whether it works (and of course if it doesn't I can
still try sudo apt-get remove python3.7 and change the symlinks
/usr/bin/python and /usr/bin/python3 to my distribution's standard
python3.5).

Thank you for your help!
Best regards,
Simon

[Matthias Köppe]

On Monday, March 30, 2020 at 8:59:06 PM UTC-4, Simon King wrote:

On 2020-03-31, Matthias Köppe <matthia...@gmail.com> wrote:
> On Monday, March 30, 2020 at 8:30:34 PM UTC-4, Simon King wrote:
>>
>> I installed python3 after "sudo add-apt-repository ppa:deadsnakes/ppa".
>> Has that been a mistake?
>>
>>
> I don't know, but I found these instructions:
> https://launchpad.net/~deadsnakes/+archive/ubuntu/ppa
> that also mention what packages should be installed

Excellent! After installing them and doing ./configure, I see this line:
 python3-3.7.3.p1:                            using system package; SPKG will not be installed

Great, this looks like progress. And thanks for telling me about the deadsnakes ppa, this may be useful at some point.

 

So, now I can try whether it works

If it doesn't, please post the logs, as usual.

Best,

Matthias

 

[Dima Pasechnik]

I'm getting this silly error while installing pillow:

Copying package files from temporary location
/home/scratch2/dimpase/sage/sage/local/var/tmp/sage/build/pillow-5.3.0.p0/inst
to /home/scratch2/dimpase/sage/sage/local
cp: cannot overwrite non-directory
'/home/scratch2/dimpase/sage/sage/local/./lib64' with directory
'/home/scratch2/dimpase/sage/sage/local/var/tmp/sage/build/pillow-5.3.0.p0/inst/home/scratch2/dimpase/sage/sage/local/./lib64'
************************************************************************
Error copying files for pillow-5.3.0.p0.

--------------

note that
$ ls -l /home/scratch2/dimpase/sage/sage/local/./lib64
lrwxrwxrwx 1 dimpase staff 3 Mar 30 18:26
/home/scratch2/dimpase/sage/sage/local/./lib64 -> lib

On Sun, Mar 29, 2020 at 8:59 PM Volker Braun <vbrau...@gmail.com> wrote:
>
> As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html
>

> I don't see any major issues for 9.1 any more, so we should soon have the first rc. If you have anything that you thing should go in then now is the time to finish it up ;-)
>
>
> 6a4580546f (tag: 9.1.beta9, trac/develop) Updated SageMath version to 9.1.beta9

> 437bbd8e43 Trac #28671: fix doctests when DOT_SAGE is symbolic link
> 6f95bfaa3c Trac #29227: Plotting single text3d results in empty scene with Three.js viewer
> 9591a58d8a Trac #29399: Put Laurent polynomial/series rings into the category of algebras
> 23bd1c7112 Trac #29397: Add more cygwin system packages
> cfd5769395 Trac #29396: Discriminant of multivariate polynomial over RR or RDF gives an error
> efb4d84db1 Trac #29394: Drop the python2 patch sys_path_security-issue_16202
> 1b175bca33 Trac #29357: Add dependencies that should trigger reconfiguring
> eecd68a1a3 Trac #29354: tox.ini: Add slackware environments
> d4024bd8fc Trac #29304: tox.ini: Add docker-linuxmint
> 95ced5fb8a Trac #29301: sage-spkg: Add an option -w for warning only if spkg-check fails
> b82db0d66b Trac #29295: Add GitHub CI workflow for cygwin
> d3772bce89 Trac #29132: Segfault in NTL::ZZ_pContext::restore during exit
> eb65fd7999 Trac #29113: Reimplement `sage -i SPKG` for optional/experimental packages as `configure --enable-SPKG && make build`
> 1f44f2204c Trac #29104: Add to tox.ini some homebrew toxenvs (with a fresh install not in /usr/local), add homebrew packages
> 7dc258767b Trac #27824: spkg-configure.m4 for python3
> be770ceae1 Trac #27247: modular_decomposition should be able to return a tree
> f32a397c64 Trac #25358: Test safe directory without python subprocess
> 322fa860fa Trac #20469: Implement Ariki-Koike algebras
> d7db290643 Trac #29082: Do not stage .pc files in src/lib/, clean old generated *.pc files at 'make distclean', fix 'permission denied' errors
> d1aa8f5a30 Trac #29383: Remove obsolete scripts from src/bin/
> d4609c12f7 Trac #29376: some details in hyperelliptic curves
> 50bbc41733 Trac #29362: build/bin/sage-pip-install: Delegate to build/bin/sage-python23
> 2589227e16 Trac #28917: sagemath should not use math names with underscore when generating Latex
> d2405f2963 Trac #28052: Incorporating a function for Tautological Implication in the Logic Module
> 12abc7cd3f Trac #29381: Move testc*.sh from src/bin to build/bin
> f98317329b Trac #29380: Update e-antic to 0.1.5
> 9399c8688b Trac #29375: Clean up some _element_constructor_() methods
> 89a1805119 Trac #29368: remove files related to SageNB and deprecated function about temporary files
> fa7f062dc8 Trac #29361: openblas spkg-configure.m4: Fix the check for lapack/cblas functions
> d26d2ae5b3 Trac #29352: more system pkgs info for Debian and Fedora
> 65914bbfb5 Trac #29342: pari spkg-configure.m4 should check gphelp more carefully
> bb0b6ae990 Trac #29323: Set up cross polytope with both Vrep and Hrep (if backend supports it)
> 084838dc2b Trac #29090: sage-system-python fixup
> 675d301be6 Trac #23769: src/bin/sage-starts should be moved to build/bin
> 28957f3f6a Trac #29287: SPKG type: Make "normal/script/pip" orthogonal to "base/standard/optional/experimental"
> 22ea385799 Trac #29102: fflas_ffpack installation error on new CPUs with old compilers
> 3f19c4a831 Trac #29170: Fix R installation errors related to gfortran
> 557e19a2fc Trac #29353: fix parent of q-Catalan numbers
> c480201572 Trac #29335: Fix Singular configure so it accepts NTL installed in nonstandard locations
> be1e22c441 (tag: 9.1.beta8) Updated SageMath version to 9.1.beta8

>
> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/689eb595-2e43-49b6-ad5b-387fdac9419e%40googlegroups.com.

[Dima Pasechnik]

On Tuesday, March 31, 2020 at 10:30:39 AM UTC+8, Dima Pasechnik wrote:

I'm getting this silly error while installing pillow:

Copying package files from temporary location
/home/scratch2/dimpase/sage/sage/local/var/tmp/sage/build/pillow-5.3.0.p0/inst
to /home/scratch2/dimpase/sage/sage/local
cp: cannot overwrite non-directory
'/home/scratch2/dimpase/sage/sage/local/./lib64' with directory
'/home/scratch2/dimpase/sage/sage/local/var/tmp/sage/build/pillow-5.3.0.p0/inst/home/scratch2/dimpase/sage/sage/local/./lib64'
************************************************************************
Error copying files for pillow-5.3.0.p0.

--------------

note that
$ ls -l /home/scratch2/dimpase/sage/sage/local/./lib64
lrwxrwxrwx 1 dimpase staff 3 Mar 30 18:26
/home/scratch2/dimpase/sage/sage/local/./lib64 -> lib

this is with system Python 3 on Fedora 30 (with Sage's built Python3 everything works, so apparently there is a sneaked in lib64

path that breaks things, and I don't really know how to fix this.

These sorts of errors appeared before in Sage, cf. e.g.

https://trac.sagemath.org/ticket/26996

 

> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release+unsubscribe@googlegroups.com.

[Matthias Köppe]

On Monday, March 30, 2020 at 11:09:06 PM UTC-4, Dima Pasechnik wrote:

On Tuesday, March 31, 2020 at 10:30:39 AM UTC+8, Dima Pasechnik wrote:

I'm getting this silly error while installing pillow:

Copying package files from temporary location
/home/scratch2/dimpase/sage/sage/local/var/tmp/sage/build/pillow-5.3.0.p0/inst
to /home/scratch2/dimpase/sage/sage/local
cp: cannot overwrite non-directory
'/home/scratch2/dimpase/sage/sage/local/./lib64' with directory
'/home/scratch2/dimpase/sage/sage/local/var/tmp/sage/build/pillow-5.3.0.p0/inst/home/scratch2/dimpase/sage/sage/local/./lib64'
************************************************************************
Error copying files for pillow-5.3.0.p0.

--------------

note that
$ ls -l /home/scratch2/dimpase/sage/sage/local/./lib64
lrwxrwxrwx 1 dimpase staff 3 Mar 30 18:26
/home/scratch2/dimpase/sage/sage/local/./lib64 -> lib

this is with system Python 3 on Fedora 30 (with Sage's built Python3 everything works, so apparently there is a sneaked in lib64

path that breaks things, and I don't really know how to fix this.

These sorts of errors appeared before in Sage, cf. e.g.

https://trac.sagemath.org/ticket/26996

I worked around this for gfortran too -- but this really needs fixing in #26996.

[Simon King]

Hi all,

It *almost* worked. On Ubuntu 16.04 with Python 3.7.7 installed from a
ppa, I get
sage -t --warn-long 44.6 src/sage/rings/padics/padic_lattice_element.py # 3 doctests failed
sage -t --warn-long 44.6 src/sage/interfaces/psage.py # 1 doctest failed

with SageMath version 9.1.beta9, Release Date: 2020-03-29.

I posted config.log, dochtml.doc, install.log, start.log and test.log
here: https://users.fmi.uni-jena.de/~king/logs/

Best regards,
Simon

[Dima Pasechnik]

On Tue, Mar 31, 2020 at 7:46 PM Simon King <simon...@uni-jena.de> wrote:
>
> Hi all,
>
> On 2020-03-31, Matthias Köppe <matthia...@gmail.com> wrote:
> >> > I don't know, but I found these instructions:
> >> > https://launchpad.net/~deadsnakes/+archive/ubuntu/ppa
> >> > that also mention what packages should be installed
> >>
> >> Excellent! After installing them and doing ./configure, I see this line:
> >> python3-3.7.3.p1: using system package; SPKG
> >> will not be installed
> >>
> >>
> > Great, this looks like progress. And thanks for telling me about the
> > deadsnakes ppa, this may be useful at some point.
> >
> >
> >> So, now I can try whether it works
> >>
> >>
> > If it doesn't, please post the logs, as usual.
>
> It *almost* worked. On Ubuntu 16.04 with Python 3.7.7 installed from a
> ppa, I get
> sage -t --warn-long 44.6 src/sage/rings/padics/padic_lattice_element.py # 3 doctests failed

this is a well-known bug/feature - the authors of this code have hard
time deciding whether
this should remain experimental, or not...
(I don't understand why on some platforms these tests fail, while on
some they don't)
https://trac.sagemath.org/ticket/29272

> sage -t --warn-long 44.6 src/sage/interfaces/psage.py # 1 doctest failed
>
> with SageMath version 9.1.beta9, Release Date: 2020-03-29.
>
> I posted config.log, dochtml.doc, install.log, start.log and test.log
> here: https://users.fmi.uni-jena.de/~king/logs/
>
> Best regards,
> Simon
>

> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/r5vaih%24tn5%241%40ciao.gmane.io.

[Steven Trogdon]

A feature or a bug? 'make distclean && make' always uses the system python to build Sage. Is there some way to force building and using the Sage-supplied python?

[Dima Pasechnik]

On Tue, Mar 31, 2020 at 9:35 PM Steven Trogdon <steven....@gmail.com> wrote:
>
> A feature or a bug? 'make distclean && make' always uses the system python to build Sage. Is there some way to force building and using the Sage-supplied python?

sure, you can use ./configure -h to see how to have packages built
Namely

./configure --with-system-python3=no

will do this for you.

[Matthias Köppe]

You can override this by:

./configure PYTHON3=/usr/bin/python3

but then you will likely run into the issues discussed on the Python 3.8 upgrade ticket (https://trac.sagemath.org/ticket/27754).

Matthias

[Steven Trogdon]

I'm replying to my previous post.

I rebuilt Sage using the Sage-provided python instead my system python and doctested Sage. The failure in doctest/test.py disappears. However, the failures in rings/integer.pyx still persist. I suspect a system component is the culprit. Any ideas which one? I had doctested 9.1.beta8 with no failures, so something has changed.

[Sébastien Labbé]

Testing with many optional and external packages, I get

Using --optional=4ti2,cbc,ccache,cryptominisat,dot2tex,e_antic,external,glucose,latte_int,lidia,lrslib,memlimit,normaliz,notedown,openssl,pandoc_attributes,pycosat,pynormaliz,rst2ipynb,sage,sage_numerical_backends_coin,sage_numerical_backends_cplex,sage_numerical_backends_gurobi

----------------------------------------------------------------------
sage -t --long src/sage/databases/findstat.py  # 8 doctests failed
sage -t --long src/sage/databases/oeis.py  # 1 doctest failed
sage -t --long src/sage/geometry/polyhedron/base.py  # Bad exit: 14
sage -t --long src/sage/graphs/generators/smallgraphs.py  # 2 doctests failed
sage -t --long src/sage/plot/animate.py  # 7 doctests failed
sage -t --long src/sage/symbolic/integration/external.py  # 1 doctest failed
----------------------------------------------------------------------
External software detected for doctesting: cplex,ffmpeg,graphviz,gurobi,imagemagick,internet,latex,pandoc

Rerunning failed tests, I get:

----------------------------------------------------------------------
sage -t --long src/sage/databases/findstat.py  # 8 doctests failed
sage -t --long src/sage/databases/oeis.py  # 1 doctest failed
sage -t --long src/sage/geometry/polyhedron/base.py  # Bad exit: 1
sage -t --long src/sage/graphs/generators/smallgraphs.py  # 2 doctests failed
sage -t --long src/sage/plot/animate.py  # 7 doctests failed
----------------------------------------------------------------------
External software detected for doctesting: ffmpeg,imagemagick,internet

These are known issues. Follow up at https://trac.sagemath.org/ticket/25536

The external.py failing test which does not reproduce is copied below.

sage -t --long src/sage/symbolic/integration/external.py
**********************************************************************
File "src/sage/symbolic/integration/external.py", line 91, in sage.symbolic.integration.external.mma_free_integrator
Failed example:
    integral(sin(y)^2, y, algorithm='mathematica_free') # optional - internet
Expected:
    1/2*y - 1/4*sin(2*y)
Got:
    -1/2*cos(y)*sin(y) + 1/2*y
**********************************************************************
1 item had failures:
   1 of   7 in sage.symbolic.integration.external.mma_free_integrator
    8 fricas tests not run
    0 tests not run because we ran out of time
    [36 tests, 1 failure, 21.85 s]

[Markus Wageringel]

I also see the error in rings/integer.pyx again now. This is https://trac.sagemath.org/ticket/28106. The doctests are limited to 3300 MB of memory be default, but this test exceeds it. The limit can be manually increased using the `--memlimit` option.

[Markus Wageringel]

Since this beta, I get the following permanent failure:

sage -t --long src/sage/env.py
**********************************************************************
File "src/sage/env.py", line 15, in sage.env
Failed example:
    out == repr((SAGE_ROOT, SAGE_LOCAL))                                        # long time
Expected:
    True
Got:
    False

Here, SAGE_ROOT and SAGE_LOCAL look correct, but the `out` variable from the preceding tests is

    sage: env = {k:v for (k,v) in os.environ.items() if not k.startswith("SAGE_")}
    sage: from subprocess import check_output
    sage: cmd = "from sage.all import SAGE_ROOT, SAGE_LOCAL; print((SAGE_ROOT, SAGE_LOCAL))"
    sage: out = check_output([sys.executable, "-c", cmd], env=env).decode().strip()   # long time
    sage: out
    "('/home/math/sagebot/sage', '/amd/compute/sagebot/sage/local')"

The left path is actually a symlink in my home directory to my sage root directory on a different partition:

    sage: SAGE_ROOT
    '/amd/compute/sagebot/sage'

Moreover, my patchbot failed multiple times with the following possibly related error:

    ...
    Sage build/upgrade complete!
    make[1]: Verzeichnis „/amd/compute/sagebot/sage“ wird verlassen
    sh: /amd/compute/sagebot/sage/local/bin/sage-starts: Datei oder Verzeichnis nicht gefunden

That is, the file `local/bin/sage-starts` is not found. This is with CentOS 7.7.

[Matthias Köppe]

On Tuesday, March 31, 2020 at 4:18:08 PM UTC-4, Markus Wageringel wrote:

 my patchbot failed multiple times with the following possibly related error:

    ...
    Sage build/upgrade complete!
    make[1]: Verzeichnis „/amd/compute/sagebot/sage“ wird verlassen
    sh: /amd/compute/sagebot/sage/local/bin/sage-starts: Datei oder Verzeichnis nicht gefunden

That is, the file `local/bin/sage-starts` is not found. This is with CentOS 7.7.

This script was moved in https://trac.sagemath.org/ticket/23769

[Matthias Köppe]

On Tuesday, March 31, 2020 at 4:18:08 PM UTC-4, Markus Wageringel wrote:

Since this beta, I get the following permanent failure:

sage -t --long src/sage/env.py
**********************************************************************
File "src/sage/env.py", line 15, in sage.env
Failed example:
    out == repr((SAGE_ROOT, SAGE_LOCAL))                                        # long time
Expected:
    True
Got:
    False

Here, SAGE_ROOT and SAGE_LOCAL look correct, but the `out` variable from the preceding tests is

    sage: env = {k:v for (k,v) in os.environ.items() if not k.startswith("SAGE_")}
    sage: from subprocess import check_output
    sage: cmd = "from sage.all import SAGE_ROOT, SAGE_LOCAL; print((SAGE_ROOT, SAGE_LOCAL))"
    sage: out = check_output([sys.executable, "-c", cmd], env=env).decode().strip()   # long time
    sage: out
    "('/home/math/sagebot/sage', '/amd/compute/sagebot/sage/local')"

The left path is actually a symlink in my home directory to my sage root directory on a different partition:

    sage: SAGE_ROOT
    '/amd/compute/sagebot/sage'

Looks like this new doctest (from https://trac.sagemath.org/ticket/29287) needs some os.path.abspath in it. Please open a ticket for this.

 

[Matthias Köppe]

On Tuesday, March 31, 2020 at 1:40:19 PM UTC-4, Steven Trogdon wrote:

I rebuilt Sage using the Sage-provided python instead my system python and doctested Sage. The failure in doctest/test.py disappears. However, the failures in rings/integer.pyx still persist. I suspect a system component is the culprit. Any ideas which one? I had doctested 9.1.beta8 with no failures, so something has changed.

Sage is able to use more system packages compared to 9.1.beta8 (via the spkg-configure.m4 mechanism). I recommend to compare config.log with the one in the earlier build, checking messages regarding the system packages.

[John H Palmieri]

Some good news and some weird news: in my attempts to debug this, I uninstalled a bunch of my homebrew packages, and I got Sage to build and pass tests. Yay. Here's the strange part: if I add the package libpng, then Sage doesn't build. With these packages, it builds and passes tests (output from "brew list --versions"):

autoconf 2.69
automake 1.16.2
boost 1.72.0_1
cabextract 1.9.1
emacs-mac emacs-26.3-z-mac-7.9
gcc 9.2.0_1 9.3.0
gdbm 1.18.1
gettext 0.20.1
gmp 6.2.0 6.1.2_2
gnutls 3.6.12
icu4c 64.2
isl 0.22.1 0.21 0.18
jpeg 9d
libdvdcss 1.4.2
libevent 2.1.11_1
libffi 3.2.1
libidn2 2.3.0
libmpc 1.1.0
libtasn1 4.14 4.16.0
libtiff 4.1.0
libunistring 0.9.10
mpfr 4.0.2
nettle 3.4.1
openssl@1.1 1.1.1d
p11-kit 0.23.20 0.23.18.1
pkg-config 0.29.2
readline 8.0.4
sqlite 3.31.1
unbound 1.9.4 1.10.0
xz 5.2.5

If I then do "brew install libpng" and build Sage from scratch, kiwisolver fails to build. For some reason, something in the build process thinks that "-sdk" is a good flag to add to "g++":

    g++ -sdk macosx clang -bundle -undefined dynamic_lookup -L/Users/palmieri/Desktop/Sage_stuff/sage_builds/TESTING/png/sage-9.1.beta9/local/lib -Wl,-rpath,/Users/palmieri/Desktop/Sage_stuff/sage_builds/TESTING/png/sage-9.1.beta9/local/lib build/temp.macosx-10.14-x86_64-3.7/py/kiwisolver.o build/temp.macosx-10.14-x86_64-3.7/py/constraint.o build/temp.macosx-10.14-x86_64-3.7/py/expression.o build/temp.macosx-10.14-x86_64-3.7/py/solver.o build/temp.macosx-10.14-x86_64-3.7/py/strength.o build/temp.macosx-10.14-x86_64-3.7/py/term.o build/temp.macosx-10.14-x86_64-3.7/py/variable.o -o build/lib.macosx-10.14-x86_64-3.7/kiwisolver.cpython-37m-darwin.so
    clang: error: unknown argument: '-sdk'
    clang: error: no such file or directory: 'macosx'
    clang: error: no such file or directory: 'clang'
    error: command 'g++' failed with exit status 1

I didn't see this before, so if I add enough other packages, this problem will somehow resolve itself. This is specific to 9.1.beta9 (i.e., using the system python3): kiwisolver builds with 9.1.beta8 and these homebrew packages.

[Matthias Köppe]

On Tuesday, March 31, 2020 at 8:02:48 PM UTC-4, John H Palmieri wrote:

[...] 

kiwisolver fails to build. For some reason, something in the build process thinks that "-sdk" is a good flag to add to "g++":

    g++ -sdk macosx clang -bundle -undefined dynamic_lookup -L/Users/palmieri/Desktop/Sage_stuff/sage_builds/TESTING/png/sage-9.1.beta9/local/lib -Wl,-rpath,/Users/palmieri/Desktop/Sage_stuff/sage_builds/TESTING/png/sage-9.1.beta9/local/lib build/temp.macosx-10.14-x86_64-3.7/py/kiwisolver.o build/temp.macosx-10.14-x86_64-3.7/py/constraint.o build/temp.macosx-10.14-x86_64-3.7/py/expression.o build/temp.macosx-10.14-x86_64-3.7/py/solver.o build/temp.macosx-10.14-x86_64-3.7/py/strength.o build/temp.macosx-10.14-x86_64-3.7/py/term.o build/temp.macosx-10.14-x86_64-3.7/py/variable.o -o build/lib.macosx-10.14-x86_64-3.7/kiwisolver.cpython-37m-darwin.so
    clang: error: unknown argument: '-sdk'
    clang: error: no such file or directory: 'macosx'
    clang: error: no such file or directory: 'clang'
    error: command 'g++' failed with exit status 1

I didn't see this before, so if I add enough other packages, this problem will somehow resolve itself. This is specific to 9.1.beta9 (i.e., using the system python3): kiwisolver builds with 9.1.beta8 and these homebrew packages.

This is https://trac.sagemath.org/ticket/29408

 

[Matthias Köppe]

python3's spkg-configure refuses to use system python if we install any of the following from spkg:

sqlite libpng bzip2 xz libffi

bzip2 comes from /usr/bin

The others from homebrew.

When you install libpng in addition to the list of packages, the set of prerequisites for accepting system python3 is complete, and then the unresolved problems with the build (kiwisolver, ...) appear when system python 3 is /usr/bin/python3 (XCode's python3).

It does not happen if you have homebrew python3 in PATH. 

[John H Palmieri]

Right, I realized that, too. However, I initially had even more homebrew packages installed, Sage built using the system Python, and the build succeeded. That's when I ran into the doctest failures, which I am now trying to reproduce: trying to see which homebrew packages trigger them.

[Markus Wageringel]

Am 31.03.2020 um 23:30 schrieb Matthias Köppe <matthia...@gmail.com>:

Looks like this new doctest (from https://trac.sagemath.org/ticket/29287) needs some os.path.abspath in it. Please open a ticket for this.

Ok, I have opened https://trac.sagemath.org/ticket/29446 for this. I had not realized this was a new doctest.

[Markus Wageringel]

Am 31.03.2020 um 23:25 schrieb Matthias Köppe <matthia...@gmail.com>:

That is, the file `local/bin/sage-starts` is not found. This is with CentOS 7.7.

This script was moved in https://trac.sagemath.org/ticket/23769

It looks like the patchbot needs to be updated for this. I will send a pull request, shortly.

[John H Palmieri]

To summarize:

- if I don't install libpng (for example), then Sage builds its own Python and everything works, all tests pass.

- if I don't install Homebrew's python but enough other packages, then the OS X Python is used, and kiwisolver fails to build. Trac 29408 appears to help with that; I'm trying a full build now.

- if I install Homebrew's python but not pkg-config, then matplotlib fails to build because it can't find libpng. (The top-level configure script finds it and so decides not to build it.)

- if I also install pkg-config, the build succeeds but I get all of the doctest failures I reported before. In this last case, here's what I've installed — I tried to keep the list as small as possible:

gcc 9.2.0_1 9.3.0
gdbm 1.18.1

gmp 6.2.0 6.1.2_2
isl 0.22.1 0.21 0.18
libmpc 1.1.0
libpng 1.6.37
libunistring 0.9.10
mpfr 4.0.2
openssl@1.1 1.1.1d
pkg-config 0.29.2
python 3.7.7
readline 8.0.4
sqlite 3.31.1
xz 5.2.5

Overall, this is a precarious situation.

By the way, this is with OS X 10.15.4 and the latest Xcode,

% gcc --version
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include/c++/4.2.1

Apple clang version 11.0.3 (clang-1103.0.32.29)

Target: x86_64-apple-darwin19.4.0

[John H Palmieri]

With trac 29404 + 29408, the build succeeds with the system python, but I get many of the doctest failures that arise with Homebrew's python.

[Matthias Köppe]

Another variation to try is to use " . .homebrew-build-env "
before the build, which sets some environment variables so that homebrew's "keg-only packages" are found.  

Among other things, this affects libpng via zlib.

[John H Palmieri]

On Wednesday, April 1, 2020 at 3:02:49 PM UTC-7, Matthias Köppe wrote:

Another variation to try is to use " . .homebrew-build-env "
before the build, which sets some environment variables so that homebrew's "keg-only packages" are found.  

Among other things, this affects libpng via zlib.

I think that when homebrew is detected, the "hints" at the end of ./configure should recommend "source .homebrew-build-env" or similar. See trac #29504.

  John