Sage 9.3.beta8 released

191 views
Skip to first unread message

Volker Braun

unread,
Mar 7, 2021, 4:35:32 PM3/7/21
to sage-release
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 


fbca269f62 (HEAD -> develop, tag: 9.3.beta8, trac/develop) Updated SageMath version to 9.3.beta8
d6cf85bd50 Trac #31182: Mutability class and pickling
7dfb05af9e Trac #31180: Add snappy as a pip package
ca1e4a13a4 Trac #31177: Add graphviz, pygraphviz packages
31ec40016f Trac #31114: Add one file build/pkgs/*/repology.txt per spkg
5ba3c41d0d Trac #31036: fixdoctests fails on multiline tests
7217db8510 Trac #31003: Add minimal pytest configuration
c462eeb3a2 Trac #30984: sage_setup: Replace imports from distutils by setuptools
f0ab074e25 Trac #30892: Characteristic polynomials over complete discrete valuation rings/fields
4415f75238 Trac #30809: interactive_simplex_method: Use of \color breaks typesetting of dictionaries in the Jupyter notebook
c8bc428f89 Trac #30728: Drawing generalizations of generalized Petersen graphs
8ff8b0b90c Trac #30537: Upgrade giac to 1.6.0-47
1d5df64577 Trac #30186: containment check for standard bracketed lyndon words does not work
d6122cd297 Trac #29632: segfault in `is_identity` morphism check over number fields
73cedf7005 Trac #31045: pip, wheel, setuptools: Make wheels available in SAGE_SPKG_WHEELS
a7f0295eab Trac #30688: Crash when expanding square root of -x
61c5e093c2 Trac #30446: Update pynac to 0.7.27 to fix wrong result on symbolic exponentiation
034ca5e1b0 Trac #18416: UniqueRepresentation issue with PowerSeriesRing
08ae587ef9 Trac #30482: Remove src/sage/symbolic/todo.txt
4ceb8ac45b Trac #30426: crt with ideals
31073aa13a Trac #30174: Trivial coercion Modules(QQ)(FiniteRankFreeModule(QQ, 3)) fails
bcc6cf26f5 Trac #30000: Implement a Q-curve test for elliptic curves over number fields
e976bd8767 Trac #29497: package symengine and its Python interface
b2f9f25ec2 Trac #28143: Minimal kernel bases
da639c3ea3 Trac #25934: Problem displaying a number field (fractional) ideal
826e5c1fc7 Trac #17684: density_plot() is broken with functions involving symbolic expressions
e7399bf8ec Trac #8972: Inversion and fraction fields for power series rings
df91b09b67 Trac #31397: Work around LaTeX bug for building pdf documentation
21c7c76aff Trac #31389: Remove clash of tdlib/ with tdlib.pyx in src/sage/graphs/graph_decompositions
4f4413737f Trac #31382: bashisms in build/bin/sage-build-env
d1b1f0567f Trac #31372: pari_jupyter is not compatible with python 3.9, breaks docker build on gitlab
ad228e2f1e Trac #31357: Fixup src/VERSION.txt added in #30912
bca796beeb Trac #31279: Add package argon2-cffi (dependency of Jupyter notebook)
aa7a2090ed Trac #31191: ubuntu-bionic-standard (python 3.6): make ptest fails with UnicodeDecodeError
10b3abca15 Trac #29152: Cygwin: Fix cliquer, giac, meataxe, rw, libbraiding to build shared libraries, using `AM_LDFLAGS=-no-undefined`
dd6f37ed42 Trac #31412: GH Actions: Fix up for new linux runner default
f2fea77292 Trac #31344: homebrew: docbuild crashes, libtcl AtForkPrepare - from sage.misc.cython globals / multiprocessing
1159ab4a12 Trac #30378: (x^2).subs({x: sqrt(x)})  returns sqrt(x) instead of x
46f654efa9 Trac #29969: Make misc doctests ready for random seeds
4767972249 Trac #29967: Make interfaces doctests ready for random seeds
f44c3edaa2 Trac #29964: Make libs doctests ready for random seeds
6f07106cd4 Trac #29555: Upgrade: OpenSSL 3.0, make it a standard package
77332630b7 Trac #26563: Add doctest for elliptic integrals of the second kind
0b99fac56b Trac #26410: Make threejs / three.js viewer respect thickness in parametric_plot3d
fe431c0397 Trac #25322: Some cleanup of the libsingular polynomial constructor
d2cf1d0770 Trac #15114: disallow dangerous coercions to RIF
10f5eb7782 Trac #10332: isPositiveSemiDefinite not accessible
8ac3225c89 Trac #31064: ci-cygwin*.yml: delegate to tox, add more stages, use more specific SAGE_LOCAL
c61a5bea30 Trac #30912: sagelib: Update metadata for PyPI deployment as package sagemath-standard
94156f67ab Trac #30770: cython_aliases: Use ecl-config to determine compiler/linker flags for ecl
1a79f3f401 Trac #31227: Accept /usr/bin/python3 from XCode 12.3 on macOS 10.15 (Catalina) and 11 (Big Sur)
090547c55a Trac #31118: sage --package update-latest: Fix for package-names-containing-dashes; accept package classes :standard:, :optional: etc., restrict to normal Python packages; fix sage --package fix-checksum
8453ffb849 (tag: 9.3.beta7) Updated SageMath version to 9.3.beta7

Matthias Köppe

unread,
Mar 7, 2021, 10:28:28 PM3/7/21
to sage-release
On Sunday, March 7, 2021 at 1:35:32 PM UTC-8 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 

fbca269f62 (HEAD -> develop, tag: 9.3.beta8, trac/develop) Updated SageMath version to 9.3.beta8

Thanks, Volker, for preparing the new release.


The first automatic tests at https://github.com/sagemath/sage/runs/2051745051 have completed:
ubuntu-{trusty,xenial,bionic,focal,groovy,hirsute}-{minimal,standard}: Clean
debian-{jessie,stretch,buster,bullseye,sid}-{minimal,standard}: Clean
... with the exception of:
{debian-bullseye,ubuntu-groovy}-standard: cvxopt testsuite errors, numerics-related sage testsuite errors
(is system BLAS feeling OK??)





Some remarks on what is new in 9.3.beta8:

9.3.beta8 makes a number of new packages available that developers may want to test or start using:
 
7dfb05af9e Trac #31180: Add snappy as a pip package
ca1e4a13a4 Trac #31177: Add graphviz, pygraphviz packages
e976bd8767 Trac #29497: package symengine and its Python interface

9.3.beta8 also contains a solution to our long-standing problem with openssl:
 
6f07106cd4 Trac #29555: Upgrade: OpenSSL 3.0, make it a standard package

For developers who test on macOS: 

> f2fea77292 Trac #31344: homebrew: docbuild crashes, libtcl AtForkPrepare - from sage.misc.cython globals / multiprocessing

Thierry Dumont

unread,
Mar 8, 2021, 5:27:44 AM3/8/21
to sage-r...@googlegroups.com
Hi,

Migration from beta 7 to beta 8 (Ubuntu 20.10):

-compilation: ok.

- make ptestlong:

1 error:

sage -t --long --warn-long 89.4 --random-seed=0
src/sage/interfaces/singular.py
Killed due to segmentation fault
**********************************************************************
Tests run before process (pid=134886) failed:

sage: R1 = singular.ring(0, '(x,y)', 'dp') ## line 65 ##
sage: R1 ## line 66 ##
polynomial ring, over a field, global ordering
// coefficients: QQ
// number of vars : 2
// block 1 : ordering dp
// : names x y
// block 2 : ordering C
sage: f = singular('9x16 - 18x13y2 - 9x12y3 + 9x10y4 - 18x11y2 + 36x8y4
+ 18x7y5 - 18x5y6 + 9x6y4 - 18x3y6 - 9x2y7 + 9y8') ## line 73 ##
sage: f ## line 74 ##
9*x^16-18*x^13*y^2-9*x^12*y^3+9*x^10*y^4-18*x^11*y^2+36*x^8*y^4+18*x^7*y^5-18*x^5*y^6+9*x^6*y^4-18*x^3*y^6-9*x^2*y^7+9*y^8
sage: f.parent() ## line 76 ##
Singular
sage: F = f.factorize(); F ## line 81 ##
[1]:
_[1]=9
_[2]=x^6-2*x^3*y^2-x^2*y^3+y^4
_[3]=-x^5+y^2
[2]:
1,1,2
sage: F[1] ## line 91 ##
9,
x^6-2*x^3*y^2-x^2*y^3+y^4,
-x^5+y^2
sage: F[1][2] ## line 95 ##
x^6-2*x^3*y^2-x^2*y^3+y^4
sage: g = f.sage(); g ## line 103 #

Matthias Köppe

unread,
Mar 8, 2021, 9:16:41 AM3/8/21
to sage-release
On Sunday, March 7, 2021 at 7:28:28 PM UTC-8 Matthias Köppe wrote:
On Sunday, March 7, 2021 at 1:35:32 PM UTC-8 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 

fbca269f62 (HEAD -> develop, tag: 9.3.beta8, trac/develop) Updated SageMath version to 9.3.beta8

Thanks, Volker, for preparing the new release.

The first automatic tests at https://github.com/sagemath/sage/runs/2051745051 have completed:
ubuntu-{trusty,xenial,bionic,focal,groovy,hirsute}-{minimal,standard}: Clean
debian-{jessie,stretch,buster,bullseye,sid}-{minimal,standard}: Clean
... with the exception of:
{debian-bullseye,ubuntu-groovy}-standard: cvxopt testsuite errors, numerics-related sage testsuite errors
(is system BLAS feeling OK??)


More results:

linuxmint-17-{minimal,standard}: Doctest errors from "setlocale" output
linuxmint-{18,19,19.3,20.1}-{minimal,standard}: Clean

fedora-26-minimal, centos-7-minimal, centos-7-i386-minimal: openssl problem, fixed by https://trac.sagemath.org/ticket/31439
fedora-{26,27,29,30,31,32,33}-standard: Clean
fedora-{27,28,29,30,31,32,33}-minimal: Clean
fedora-28-standard: 
- cvxopt testsuite errors
- src/sage/rings/polynomial/groebner_fan.py # 3 doctests failed
- src/sage/rings/real_mpfr.pyx # Killed due to abort
fedora-24-standard:
  [eclib-20190909]   /usr/bin/ld: /usr/lib64/libntl.so: undefined reference to `std::__exception_ptr::exception_ptr::_M_release()@CXXABI_1.3.13'

centos-7-standard: [dochtml] ImportError: /lib64/libz.so.1: version `ZLIB_1.2.9' not found (required by /sage/local/lib/libpng16.so.16)
centos-8-{minimal,standard}: Clean

gentoo-{minimal,standard}: Clean
gentoo-python3.7-{minimal,standard}: Clean

archlinux-latest-{minimal,standard}: Clean

slackware-14.2-{minimal,standard}: Clean

conda-forge-minimal: Error building package python3-3.9.1
conda-forge-standard: Sage does not accept system python3 - fixed by https://trac.sagemath.org/ticket/31358?

ubuntu-{bionic,focal}-i386-minimal: [openssl-3.0.0-alpha11] /usr/include/stdio.h:27:10: fatal error: bits/libc-header-start.h: No such file or directory
debian-buster-i386-minimal: likewise
debian-buster-i386-standard: Clean except for
- src/sage/parallel/decorate.py # 1 doctest failed
- src/sage/schemes/elliptic_curves/ell_field.py # 1 doctest failed
 


Emmanuel Charpentier

unread,
Mar 8, 2021, 10:59:11 AM3/8/21
to sage-release

FWIW, on Debian testing running on core i7 + 16 GB RAM, upgrading 9.3.beta7 (configured to use all available system packages) to 9.3.beta8 leads to :

  • ./configure still insists on installing a bunch of already installed and up-to-date system packages.

  • ptestlong results in two permanent and one transient failures, the former already seen and reported in previous versions :

File Results P/T
src/sage/tests/parigp.py Timed out T
src/sage/tests/gap_packages.py 1 doctest failed P
src/sage/groups/abelian_gps/abelian_group_gap.py 1 doctest failed P
​

HTH,

François Bissey

unread,
Mar 8, 2021, 2:52:46 PM3/8/21
to 'Justin C. Walker' via sage-release
I have been having a segfault in that file for a while between 9.3.beta7 and 9.3.beta8
in sage-on-gentoo when running parallel testing. However the tests run all fine on that
file when doing an individual run. There probably can be some kind of race condition
when running in parallel.

Matthias Köppe

unread,
Mar 8, 2021, 3:01:21 PM3/8/21
to sage-release
See https://trac.sagemath.org/ticket/30945

François Bissey

unread,
Mar 8, 2021, 3:03:07 PM3/8/21
to sage-r...@googlegroups.com
Exactly that, thanks for adding me. I can reliably reproduce it.
> --
> 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/f0a6f03c-cbf4-4a88-8929-8dd453c6c9e7n%40googlegroups.com.

Sébastien Labbé

unread,
Mar 9, 2021, 11:19:29 AM3/9/21
to sage-release
On Ubuntu 18.04, with few optional/external packages, I get:

Using --optional=4ti2,build,cbc,ccache,cryptominisat,database_jones_numfield,debian,dot2tex,e_antic,external,fricas,glucose,latte_int,libnauty,lidia,lrslib,memlimit,normaliz,notedown,pandoc_attributes,pip,pycosat,pynormaliz,rst2ipynb,sage,sage_numerical_backends_coin,sage_spkg

----------------------------------------------------------------------
sage -t --long --warn-long 72.3 --random-seed=0 src/sage/tests/cmdline.py  # 1 doctest failed
----------------------------------------------------------------------
External software detected for doctesting: ffmpeg,graphviz,imagemagick,internet,latex,octave,pandoc

Rerunning failed tests, I get:

----------------------------------------------------------------------
sage -t --long --warn-long 74.8 --random-seed=0 src/sage/tests/cmdline.py  # 1 doctest failed
----------------------------------------------------------------------
External software detected for doctesting: internet,pandoc


The failing doctest is:


sage -t --long --warn-long 74.8 --random-seed=0 src/sage/tests/cmdline.py
**********************************************************************
File "src/sage/tests/cmdline.py", line 606, in sage.tests.cmdline.test_executable
Failed example:
    err  # optional - internet
Expected:
    ''
Got:
    '/home/slabbe/GitBox/sage/local/lib/python3.8/site-packages/sage/misc/package.py:114: UserWarning: failed to fetch the version of pkg=\'ore_algebra\' at https://pypi.org/pypi/ore_algebra/json\n  warnings.warn("failed to fetch the version of pkg={!r} at {}".format(pkg, url))\n'
**********************************************************************
1 item had failures:
   1 of 260 in sage.tests.cmdline.test_executable
    3 gdb tests not run
    6 py2 tests not run
    6 sagenb tests not run
    0 tests not run because we ran out of time
    [259 tests, 1 failure, 29.70 s]


Everything else is okay! Congratulation. And yes thank you Volker for the
release.

Justin C. Walker

unread,
Mar 9, 2021, 3:14:15 PM3/9/21
to sage-r...@googlegroups.com


> On Mar 7, 2021, at 13:35 , 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 successfully on two macOS platforms ( 10.13.6, 2017 iMac, 18-core Xeon W; 10.14.6 2017 MBP, Core i7).
Testing (‘ptestlong’) had no failures on either one.

Build/test on a 2019 MBP, macOS 10.15.7 (8-core Core i9) will occur later this week.

Justin

PS: no package managers were used or harmed in this process.

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



Justin C. Walker

unread,
Mar 11, 2021, 12:18:31 AM3/11/21
to 'jonatha...@googlemail.com' via sage-release


> On Mar 9, 2021, at 12:14, 'Justin C. Walker' via sage-release <sage-r...@googlegroups.com> wrote:
>
>
>
>> On Mar 7, 2021, at 13:35 , 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 successfully on two macOS platforms ( 10.13.6, 2017 iMac, 18-core Xeon W; 10.14.6 2017 MBP, Core i7).
> Testing (‘ptestlong’) had no failures on either one.
>
> Build/test on a 2019 MBP, macOS 10.15.7 (8-core Core i9) will occur later this week.

Finally, I am able to build a 9.3 beta on 10.15 macOS system. No problems with the build, and all tests passed!

Thanks for the hard work to get this working. Not sure what made the difference for “make configure”. In fact, now that works for previous betas.

Justin


Samuel Lelièvre

unread,
Mar 12, 2021, 4:00:17 PM3/12/21
to sage-release
Thanks for this new release.

On macOS 10.14.6, with Homebrew, failed to build cysignals-1.10.2.

Logs:

https://www.math.u-psud.fr/~lelievre/t/slel-2021-03-12-s93b8.zip

Matthias Köppe

unread,
Mar 13, 2021, 12:27:35 PM3/13/21
to sage-release
From your log:

  gcc -bundle -undefined dynamic_lookup -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX10.14.sdk -L/opt/s/sage93c/local/lib -Wl,-rpath,/opt/s/sage93c/local/lib -O2 -g -march=native -Wp,-U_FORTIFY_SOURCE build/temp.macosx-10.14-x86_64-3.9/build/src/cysignals/signals.o -L/usr/local/lib -L/usr/local/opt/openssl@1.1/lib -L/usr/local/opt/sqlite/lib -L/usr/local/opt/tcl-tk/lib -o build/lib.macosx-10.14-x86_64-3.9/cysignals/signals.cpython-39-darwin.so -lpari -pthread
  ld: illegal data reference in _PARI_SIGINT_block to thread local variable _PARI_SIGINT_block in dylib /opt/s/sage93c/local/lib/libpari.dylib for architecture x86_64
  clang: error: linker command failed with exit code 1 (use -v to see invocation)
  error: command '/usr/bin/gcc' failed with exit code 1
  Building wheel for cysignals (setup.py): finished with status 'error'
 
This is another instance of homebrew packages leaking into the build, in this case pari.

Please try with https://trac.sagemath.org/ticket/31335 ("homebrew: Unused packages (singular, pari, ...) in /usr/local leak into sagelib build via distutils.cfg")






Matthias Köppe

unread,
Mar 13, 2021, 2:21:20 PM3/13/21
to sage-release
Results for macOS 10.15 (Catalina) are now in at https://github.com/sagemath/sage/runs/2051746258

homebrew-macos-10.15-minimal-xcode_11.7 (https://github.com/sagemath/sage/runs/2051746258): Clean except for:
sage -t --random-seed=0 src/sage/geometry/polyhedron/base.py  # 1 doctest failed (get_memory_usage)
sage -t --random-seed=0 src/sage/interfaces/gap.py  # 1 doctest failed
(File "src/sage/interfaces/gap.py", line 482, in sage.interfaces.gap.Gap_generic._read_in_file_command)

-standard-default likewise
-standard-python3_xcode (https://github.com/sagemath/sage/runs/2051746451) likewise 
-minimal-nokegonly-xcode_11.7 (https://github.com/sagemath/sage/runs/2051746494) likewise

Lots of additional doctest failures from unexpected linker messages:
File "src/sage/structure/factory.pyx", line 199, in sage.structure.factory.UniqueFactory
Failed example:
    cython("cdef class C: pass")
Expected nothing
Got:
    ld: warning: dylib (/usr/local/lib/libmpfr.dylib) was built for newer macOS version (10.15) than being linked (10.9)
...-
-standard-python3_xcode likewise
-minimal-nokegonly-{xcode_12.2,default,xcode_12.4} likewise

homebrew-macos-10.15-standard-nokegonly-python3_xcode-{xcode_11.7,xcode_12.2,default,xcode_12.4} (https://github.com/sagemath/sage/runs/2051746530): fpylll build fails (does not find fplll in /usr/local)


There are no results for macOS 11 (Big Sur) -- we will have to make adjustments to the CI workflow so that they get run. This is https://trac.sagemath.org/ticket/31492



 

Samuel Lelièvre

unread,
Mar 14, 2021, 1:48:30 PM3/14/21
to sage-release
2021-03-13 17:27 UTC, Matthias Köppe:
>
> On Friday, March 12, 2021 at 1:00:17 PM UTC-8 Samuel Lelievre wrote:
>>
>> On macOS 10.14.6, with Homebrew, failed to build cysignals-1.10.2.
>>
>> Logs:
>>
>> https://www.math.u-psud.fr/~lelievre/t/slel-2021-03-12-s93b8.zip
>
> From your log:
>
> gcc -bundle -undefined dynamic_lookup -isysroot /Library/Developer/CommandLineTools/SDKs/MacOSX10.14.sdk -L/opt/s/sage93c/local/lib -Wl,-rpath,/opt/s/sage93c/local/lib -O2 -g -march=native -Wp,-U_FORTIFY_SOURCE build/temp.macosx-10.14-x86_64-3.9/build/src/cysignals/signals.o -L/usr/local/lib -L/usr/local/opt/openssl@1.1/lib -L/usr/local/opt/sqlite/lib -L/usr/local/opt/tcl-tk/lib -o build/lib.macosx-10.14-x86_64-3.9/cysignals/signals.cpython-39-darwin.so -lpari -pthread
> ld: illegal data reference in _PARI_SIGINT_block to thread local variable _PARI_SIGINT_block in dylib /opt/s/sage93c/local/lib/libpari.dylib for architecture x86_64
> clang: error: linker command failed with exit code 1 (use -v to see invocation)
> error: command '/usr/bin/gcc' failed with exit code 1
> Building wheel for cysignals (setup.py): finished with status 'error'
>
> This is another instance of homebrew packages
> leaking into the build, in this case pari.
>
> Please try with https://trac.sagemath.org/ticket/31335
> ("homebrew: Unused packages (singular, pari, ...)
> in /usr/local leak into sagelib build via distutils.cfg")

Tried with #31335 applied: cysignals still fails to build.

https://www.math.u-psud.fr/~lelievre/t/slel-2021-03-14-s93b8-t31335.zip

Matthias Köppe

unread,
Mar 14, 2021, 1:53:02 PM3/14/21
to sage-release
On Sunday, March 14, 2021 at 10:48:30 AM UTC-7 Samuel Lelievre wrote:
2021-03-13 17:27 UTC, Matthias Köppe:
> On Friday, March 12, 2021 at 1:00:17 PM UTC-8 Samuel Lelievre wrote:
>> On macOS 10.14.6, with Homebrew, failed to build cysignals-1.10.2.
> This is another instance of homebrew packages
> leaking into the build, in this case pari.
>
> Please try with https://trac.sagemath.org/ticket/31335
> ("homebrew: Unused packages (singular, pari, ...)
> in /usr/local leak into sagelib build via distutils.cfg")

Tried with #31335 applied: cysignals still fails to build.

https://www.math.u-psud.fr/~lelievre/t/slel-2021-03-14-s93b8-t31335.zip

Thanks for testing!

Could you try with: export SETUPTOOLS_USE_DISTUTILS=local
before restarting build?



 

Matthias Köppe

unread,
Mar 14, 2021, 1:57:13 PM3/14/21
to sage-release







Matthias Köppe

unread,
Mar 14, 2021, 7:09:30 PM3/14/21
to sage-release
On Sunday, March 14, 2021 at 10:57:13 AM UTC-7 Matthias Köppe wrote:
Results for macOS 11 (Big Sur) using https://trac.sagemath.org/ticket/31492 at https://github.com/mkoeppe/sage/runs/2103782309

local-macos (homebrew-macos, minimal, default, macos-11.0) 


local-macos (homebrew-macos, standard, default, macos-11.0)
mostly clean (times out during make ptest)


 

Samuel Lelièvre

unread,
Mar 15, 2021, 9:00:13 PM3/15/21
to sage-release
2021-03-14 17:53 UTC, Matthias Köppe:
That trick helps.

a.

Starting from the same branch that failed above, this
```
$ source .homebrew-build-env
$ export SETUPTOOLS_USE_DISTUTILS=local
$ ./bootstrap -q && ./configure -q && make -s V=0
```
succeeds in building cysignals-1.10.2 and many more
packages but eventually fails to build argon2_cffi-20.1.0.

Logs:
https://www.math.u-psud.fr/~lelievre/t/slel-2021-03-14-s93b8-t31335-trick.zip

b.

Starting from a fresh clone, similar result.

Logs:
https://www.math.u-psud.fr/~lelievre/t/slel-2021-03-14-s93b8-t31335-trick-fresh.zip

c.

In a different install, applying tickets 31474 (upgrade cysignals)
and 31335, the build fails on distutil.

https://www.math.u-psud.fr/~lelievre/t/slel-2021-03-14-s93b8-t31474-t31335.zip

Matthias Köppe

unread,
Mar 15, 2021, 9:51:21 PM3/15/21
to sage-release
On Monday, March 15, 2021 at 6:00:13 PM UTC-7 Samuel Lelievre wrote:
2021-03-14 17:53 UTC, Matthias Köppe:
> Could you try with: export SETUPTOOLS_USE_DISTUTILS=local
> before restarting build?

That trick helps.

a.

Starting from the same branch that failed above, this
```
$ source .homebrew-build-env
$ export SETUPTOOLS_USE_DISTUTILS=local
$ ./bootstrap -q && ./configure -q && make -s V=0
```
succeeds in building cysignals-1.10.2 and many more
packages

Thanks for testing!
 
but eventually fails to build argon2_cffi-20.1.0.

Yes, this one fixed in https://trac.sagemath.org/ticket/31441 (already merged in 9.3.beta9)
 
In a different install, applying tickets 31474 (upgrade cysignals)
and 31335, the build fails on distutil.

... dateutil, actually.
 
This looks like a typical error that arises when you switch the version of python3 in an existing build.
Nothing to fix, it will go away if you run "make distclean" or possibly just "rm $SAGE_LOCAL/pyvenv.cfg"


Reply all
Reply to author
Forward
0 new messages