Sage 9.8.rc0 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

9116c5584f9 (HEAD -> develop, tag: 9.8.rc0, trac/develop) Updated SageMath version to 9.8.rc0
0f25b00a8e2 Trac #34537: Upgrade to pari 2.15.2
b5c9fed6df1 Trac #34863: construct QuadraticForm and BinaryQF objects from polynomial
17b7b0ecdff Trac #34860: Uniformize headline: number theory
8fcb85e8c34 Trac #34811: rational maps for WeierstrassIsomorphism should not be polynomials
c7be32338dd Trac #34791: move ProductTree and prod_with_derivative() to sage.rings.generic
814d948988a Trac #33201: Fix "referenced before assignment" warnings in matrices
fa775740eea Trac #8598: Add graphical output to operation tables
081d67a36c1 Trac #34922: less usage of python2 long type in padics
9cff81c43f9 Trac #34921: fix E714 in pyx files inside matrix/
e52e9687b46 Trac #34919: fix E266 in modular
273d12b7460 Trac #34910: autopep8 for E3 in combinat/
0bf8cd2b73c Trac #34903: tox.ini / GH Actions: Update Linux platforms
3b249656aaa Trac #34893: fix and activate pycodestyle E211
60841095f3a Trac #34841: adapt findstat doctests to new newlines
e09c8475ae9 Trac #34771: switch to libgap in sage/coding
bd7d9ace6d1 Trac #34892: Fixing the linter (again)
54bfa669d66 Trac #34911: Adding class Kissat in sage/sat/solvers/dimacs.py
0aa67a58228 Trac #34909: create an optional package for kissat SAT solver
7cf71672816 Trac #34906: use NTL's MinPolyMod()
63ffaee72e6 Trac #34905: Improvements to _vector_ of a quantum group module
97493228e6b Trac #34904: tikz output should be 1 paragraph, to keep sagetex happy
556e1f39ffc Trac #34898: some details about min and max
20074144815 Trac #34895: fix E714 in pyx files outside of matrix/
349ce9c9da3 Trac #34870: FPLLL 5.4.4  and FPyLLL 0.5.9
62670bdf4af Trac #34858: Fix sagemath-standard sdist
83b0cabc2f2 Trac #34849: More fixes to support tox 4
1c0f253a854 Trac #34827: Adding some new functionality to the Vector Partition Module
b3abb5bb978 Trac #34809: Vizing edge coloring optimization
8d322beb1e4 Trac #34635: Add tensor_factors() method to TensorProductOfHighestWeightModules in quantum_group_gap
bf942aee422 Trac #34581: Implement substitution in InfinitePolynomialRing
88fd702d051 Trac #34447: Make EdgesView accept two sets of vertices
047281e0287 (tag: 9.8.beta7) Updated SageMath version to 9.8.beta7

[Kwankyu Lee]

Please consider #34873 for this release.

[Eric Gourgoulhon]

On Ubuntu 20.04 running on Xeon E5-2623 + 16 GB RAM:
- incremental build (-j16) from 9.8.beta7, with system python (3.8.10): OK

- make ptestlong --> all tests passed except for

sage -t --long --warn-long 66.8 --random-seed=81878619703120460056743231466438357448 src/sage/env.py
**********************************************************************
File "src/sage/env.py", line 16, in sage.env
Failed example:
    out == repr((SAGE_ROOT, SAGE_LOCAL))                                        # long time
Expected:
    True
Got:
    False
**********************************************************************
1 item had failures:
   1 of   6 in sage.env
    [45 tests, 1 failure, 7.81 s]

This is a permanent failure.

Eric.

[Emmanuel Charpentier]

On Debian testing running on core i7 + 16 GB RAM :

- upgrading 9.8.beta7 + #34935 fails ; I had to go the `make distclean ; ... ` way.

- `ptestlong` gives the same three permanent failures already reported for a while

- reinstalling `mathics` and `pari_jupyter`fail.

HTH,

[Emmanuel Charpentier]

BTW : please consider including #34935 (already positively reviewed...).

[Eric Gourgoulhon]

Running the doctests manually in a Sage console, I get

sage: out
"(None, '/home/eric/sage/9.8-develop/local/var/lib/sage/venv-python3.8')"
sage: repr((SAGE_ROOT, SAGE_LOCAL))
"('/home/eric/sage/9.8-develop', '/home/eric/sage/9.8-develop/local')"

hence the doctest failure.

On another computer with Ubuntu 22.04 (instead of 20.04), the doctest is passed,

since both out and repr((SAGE_ROOT, SAGE_LOCAL)) are

"('/home/eric/sage/9.8-develop', '/home/eric/sage/9.8-develop/local')"

Eric.

 

[Sébastien Labbé]

Thanks for the release.

On my side, I am stuck with the following. I was having the same issue with the previous version. I saw I was not alone, but I still don't know what is the solution? Anybody know? Make distclean?

Sébastien

[sagelib-9.8.rc0] installing. Log file: /home/slabbe/GitBox/sage/logs/pkgs/sagelib-9.8.rc0.log

  [sagelib-9.8.rc0] error installing, exit status 1. End of log file:
  [sagelib-9.8.rc0]             |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  [sagelib-9.8.rc0]       sage/misc/nested_class.c:1341:43: note: in expansion of macro ‘__Pyx_TRASHCAN_BEGIN_CONDITION’
  [sagelib-9.8.rc0]        1341 | #define __Pyx_TRASHCAN_BEGIN(op, dealloc) __Pyx_TRASHCAN_BEGIN_CONDITION(op,\
  [sagelib-9.8.rc0]             |                                           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  [sagelib-9.8.rc0]       sage/misc/nested_class.c:2680:3: note: in expansion of macro ‘__Pyx_TRASHCAN_BEGIN’
  [sagelib-9.8.rc0]        2680 |   __Pyx_TRASHCAN_BEGIN(o, __pyx_tp_dealloc_4sage_4misc_12nested_class_NestedClassMetaclass)
  [sagelib-9.8.rc0]             |   ^~~~~~~~~~~~~~~~~~~~
  [sagelib-9.8.rc0]       sage/misc/nested_class.c:1338:17: warning: implicit declaration of function ‘_PyTrash_thread_destroy_chain’ [-Wimplicit-function-declaration]
  [sagelib-9.8.rc0]        1338 |                 _PyTrash_thread_destroy_chain();\
  [sagelib-9.8.rc0]             |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  [sagelib-9.8.rc0]       sage/misc/nested_class.c:2683:3: note: in expansion of macro ‘__Pyx_TRASHCAN_END’
  [sagelib-9.8.rc0]        2683 |   __Pyx_TRASHCAN_END
  [sagelib-9.8.rc0]             |   ^~~~~~~~~~~~~~~~~~
  [sagelib-9.8.rc0]       error: Command "gcc-9 -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -Wno-unused -g -O2 -g -O2 -fPIC -I/home/slabbe/GitBox/sage/src -I/home/slabbe/GitBox/sage/local/var/lib/sage/venv-python3.11.1/lib/python3.11/site-packages/numpy/core/include -I/home/slabbe/GitBox/sage/local/var/lib/sage/venv-python3.11.1/include/python3.11 -I/home/slabbe/GitBox/sage/local/var/lib/sage/venv-python3.11.1/include/python3.11 -c sage/categories/category_cy_helper.c -o build/temp.linux-x86_64-cpython-311/sage/categories/category_cy_helper.o" failed with exit status 1
  [sagelib-9.8.rc0]       error: subprocess-exited-with-error
  [sagelib-9.8.rc0]      
  [sagelib-9.8.rc0]       × python setup.py develop did not run successfully.
  [sagelib-9.8.rc0]       │ exit code: 1
  [sagelib-9.8.rc0]       ╰─> See above for output.
  [sagelib-9.8.rc0]      
  [sagelib-9.8.rc0]       note: This error originates from a subprocess, and is likely not a problem with pip.
  [sagelib-9.8.rc0]       full command: /home/slabbe/GitBox/sage/local/var/lib/sage/venv-python3.11.1/bin/python3 -c '
  [sagelib-9.8.rc0]       exec(compile('"'"''"'"''"'"'
  [sagelib-9.8.rc0]       # This is <pip-setuptools-caller> -- a caller that pip uses to run setup.py
  [sagelib-9.8.rc0]       #
  [sagelib-9.8.rc0]       # - It imports setuptools before invoking setup.py, to enable projects that directly
  [sagelib-9.8.rc0]       #   import from `distutils.core` to work with newer packaging standards.
  [sagelib-9.8.rc0]       # - It provides a clear error message when setuptools is not installed.
  [sagelib-9.8.rc0]       # - It sets `sys.argv[0]` to the underlying `setup.py`, when invoking `setup.py` so
  [sagelib-9.8.rc0]       #   setuptools doesn'"'"'t think the script is `-c`. This avoids the following warning:
  [sagelib-9.8.rc0]       #     manifest_maker: standard file '"'"'-c'"'"' not found".
  [sagelib-9.8.rc0]       # - It generates a shim setup.py, for handling setup.cfg-only projects.
  [sagelib-9.8.rc0]       import os, sys, tokenize
  [sagelib-9.8.rc0]      
  [sagelib-9.8.rc0]       try:
  [sagelib-9.8.rc0]           import setuptools
  [sagelib-9.8.rc0]       except ImportError as error:
  [sagelib-9.8.rc0]           print(
  [sagelib-9.8.rc0]               "ERROR: Can not execute `setup.py` since setuptools is not available in "
  [sagelib-9.8.rc0]               "the build environment.",
  [sagelib-9.8.rc0]               file=sys.stderr,
  [sagelib-9.8.rc0]           )
  [sagelib-9.8.rc0]           sys.exit(1)
  [sagelib-9.8.rc0]      
  [sagelib-9.8.rc0]       __file__ = %r
  [sagelib-9.8.rc0]       sys.argv[0] = __file__
  [sagelib-9.8.rc0]      
  [sagelib-9.8.rc0]       if os.path.exists(__file__):
  [sagelib-9.8.rc0]           filename = __file__
  [sagelib-9.8.rc0]           with tokenize.open(__file__) as f:
  [sagelib-9.8.rc0]               setup_py_code = f.read()
  [sagelib-9.8.rc0]       else:
  [sagelib-9.8.rc0]           filename = "<auto-generated setuptools caller>"
  [sagelib-9.8.rc0]           setup_py_code = "from setuptools import setup; setup()"
  [sagelib-9.8.rc0]      
  [sagelib-9.8.rc0]       exec(compile(setup_py_code, filename, "exec"))
  [sagelib-9.8.rc0]       '"'"''"'"''"'"' % ('"'"'/home/slabbe/GitBox/sage/src/setup.py'"'"',), "<pip-setuptools-caller>", "exec"))' --no-user-cfg develop --no-deps
  [sagelib-9.8.rc0]       cwd: /home/slabbe/GitBox/sage/src/
  [sagelib-9.8.rc0]   error: subprocess-exited-with-error
  [sagelib-9.8.rc0]  
  [sagelib-9.8.rc0]   × python setup.py develop did not run successfully.
  [sagelib-9.8.rc0]   │ exit code: 1
  [sagelib-9.8.rc0]   ╰─> See above for output.
  [sagelib-9.8.rc0]  
  [sagelib-9.8.rc0]   note: This error originates from a subprocess, and is likely not a problem with pip.
  [sagelib-9.8.rc0]   *****************************************************************************************************
  [sagelib-9.8.rc0]   Error installing
  [sagelib-9.8.rc0]   *****************************************************************************************************
  [sagelib-9.8.rc0]  
  [sagelib-9.8.rc0]   real    0m12,552s
  [sagelib-9.8.rc0]   user    0m11,883s
  [sagelib-9.8.rc0]   sys    0m1,472s

  [sagelib-9.8.rc0] Full log file: /home/slabbe/GitBox/sage/logs/pkgs/sagelib-9.8.rc0.log

[Sébastien Labbé]

On another machine running Ubuntu 22.04, installing sage from source from scratch, I get "m4: Permission denied" error during the "make configure". Should I "sudo make configure" ? Never done that before...

See below:

$ make configure

./bootstrap -d
make[1] : on entre dans le répertoire « /home/slabbe/GitBox/sage »
rm -rf config/install-sh config/compile config/config.guess config/config.sub config/missing configure build/make/Makefile-auto.in
rm -f src/doc/en/installation/*.txt
rm -rf src/doc/en/reference/spkg/*.rst
rm -f environment.yml
rm -f src/environment.yml
rm -f src/environment-dev.yml
rm -f environment-optional.yml
rm -f src/environment-optional.yml
rm -f src/Pipfile
rm -f src/pyproject.toml
rm -f src/requirements.txt
rm -f src/setup.cfg
make[1] : on quitte le répertoire « /home/slabbe/GitBox/sage »
bootstrap:: installing 'm4/sage_spkg_configures.m4'
./bootstrap: installing /home/slabbe/GitBox/sage/build/pkgs/sagemath_environment/src/pyproject.toml
./bootstrap: 9: m4: Permission denied
make: *** [Makefile:365 : configure] Erreur 1

[John H Palmieri]

Some people have had an issue with Singular: the system version is too new. Using "./configure --with-system-singular=no" (probably after "make distclean"?) would help, if that's the problem. There is an upgrade ticket for Singular that has had many isses, but last time I checked, it had a positive review and was marked a blocker.

[Sébastien Labbé]

On Thursday, February 2, 2023 at 6:31:21 PM UTC+1 John H Palmieri wrote:

Some people have had an issue with Singular: the system version is too new. Using "./configure --with-system-singular=no" (probably after "make distclean"?) would help, if that's the problem. There is an upgrade ticket for Singular that has had many isses, but last time I checked, it had a positive review and was marked a blocker.

Thanks for the reply.  Well, I am experiencing the issue on the old Ubuntu 18.04 on which singular is not installed :

$ singular
singular : commande introuvable

I already tried "--with-system-singular=no" last week after reading the previous post, but I still have the same issue. I am currently trying again with rc0 and after make distclean just to make sure.

Sébastien

 

[Dima Pasechnik]

On Thu, 2 Feb 2023, 17:30 Sébastien Labbé, <sla...@gmail.com> wrote:

On another machine running Ubuntu 22.04, installing sage from source from scratch, I get "m4: Permission denied" error during the "make configure". Should I "sudo make configure" ?

no. Something it either off with permissions in SAGE_ROOT, or m4 is not installed.

Or it's a bug that slipped in due to insufficient testing on machines without installed autotools.

Never done that before...

See below:

$ make configure

./bootstrap -d
make[1] : on entre dans le répertoire « /home/slabbe/GitBox/sage »
rm -rf config/install-sh config/compile config/config.guess config/config.sub config/missing configure build/make/Makefile-auto.in
rm -f src/doc/en/installation/*.txt
rm -rf src/doc/en/reference/spkg/*.rst
rm -f environment.yml
rm -f src/environment.yml
rm -f src/environment-dev.yml
rm -f environment-optional.yml
rm -f src/environment-optional.yml
rm -f src/Pipfile
rm -f src/pyproject.toml
rm -f src/requirements.txt
rm -f src/setup.cfg
make[1] : on quitte le répertoire « /home/slabbe/GitBox/sage »
bootstrap:: installing 'm4/sage_spkg_configures.m4'
./bootstrap: installing /home/slabbe/GitBox/sage/build/pkgs/sagemath_environment/src/pyproject.toml
./bootstrap: 9: m4: Permission denied
make: *** [Makefile:365 : configure] Erreur 1

--
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/3005af48-d9e6-4399-9575-1ac642258308n%40googlegroups.com.

[Matthias Köppe]

On Thursday, February 2, 2023 at 9:30:30 AM UTC-8 Sébastien Labbé wrote:

./bootstrap: 9: m4: Permission denied

make: *** [Makefile:365 : configure] Erreur 1

Do you have "." in PATH?

What does "which m4" say?

 

[Sébastien Labbé]

On Thursday, February 2, 2023 at 7:37:27 PM UTC+1 matthia...@gmail.com wrote:

On Thursday, February 2, 2023 at 9:30:30 AM UTC-8 Sébastien Labbé wrote:

./bootstrap: 9: m4: Permission denied

make: *** [Makefile:365 : configure] Erreur 1

Do you have "." in PATH?

No

 

What does "which m4" say?

 

It returns nothing! Thanks for the hint. I found the problem: when reading the sage README to install sage from scratch on this new machine, I installed the basic stuff as follows:

$ sudo apt install gcc gfortran g++ make m4 perl ranlib tar bc

[...]

E: Impossible de trouver le paquet ranlib

But this returned an error on ranlib and I assumed it had installed the other packages in the list but no, it did not (gcc, gfortran, etc. were not installed as well).

Is ranlib really needed for sage? Can we remove it from the README?

[Emmanuel Charpentier]

In Debian, there is no package `ranlib`, the `ranlib` *binary* is included in the `binutils-<yourarch>`, which is depended upon by the pseudo-package `binutils`.

HTH,

[Dima Pasechnik]

ranlib is a standard tool, and is used by various build systems. On
Debian it's a part of binutils package.

README is not platform-specific, and it does not tell you to install
*package* called ranlib.
However, we should add there an advice to look for what packages to install in
build/pkgs/_prereq/distros/
(e.g. for Debian it's in build/pkgs/_prereq/distros/debian.txt)

(Once our GitHub is live in few days, it's a very quick PR to do :-))

>
>
>
> --
> 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/d6aecaee-279a-45f8-a645-36483d72a382n%40googlegroups.com.

[Nasser M. Abbasi]

I get this error on Linux virtual box running 

lsb_release -a
LSB Version:    n/a
Distributor ID:    ManjaroLinux
Description:    Manjaro Linux
Release:    22.0.0
Codename:    Sikaris

After extracting the gzip file, these are the commands I did

unset SAGE_ROOT 

 unset SAGE_LOCAL 

 export GIAC=/usr/local/bin/giac 

export MAXIMA=/usr/local/bin/maxima 

./configure --with-system-python=no 

make

At the end I see 

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

sagemath_doc_html-none]   File "/home/me/TMP/sage-9.8.rc0/src/sage_docbuild/sphinxbuild.py", line 263, in raise_errors
[sagemath_doc_html-none]     raise OSError(self._error)
[sagemath_doc_html-none] OSError: /home/me/TMP/sage-9.8.rc0/src/sage/manifolds/chart.py:docstring of sage.manifolds.chart.RealChart.plot:104: WARNING: Exception occurred in plotting chart-1
[sagemath_doc_html-none]
[sagemath_doc_html-none]     Note: incremental documentation builds sometimes cause spurious
[sagemath_doc_html-none]     error messages. To be certain that these are real errors, run
[sagemath_doc_html-none]     "make doc-clean doc-uninstall" first and try again.
[sagemath_doc_html-none] make[6]: *** [Makefile:28: doc-inventory--reference-manifolds] Error 1
[sagemath_doc_html-none] make[5]: *** [Makefile:45: doc-inventory-reference] Error 2
make[4]: *** [Makefile:3262: sagemath_doc_html-SAGE_DOCS-no-deps] Error 2
make[3]: *** [Makefile:3262: /home/me/TMP/sage-9.8.rc0/local/var/lib/sage/installed/sagemath_doc_html-none] Error 2
make[2]: *** [Makefile:2766: all-start] Error 2
make[2]: Leaving directory '/home/me/TMP/sage-9.8.rc0/build/make'

real    71m30.833s
user    103m7.355s
sys    5m48.306s
***************************************************************
Error building Sage.

The following package(s) may have failed to build (not necessarily
during this run of 'make all-start'):

It is safe to delete any log files and build directories, but they
contain information that is helpful for debugging build problems.
WARNING: If you now run 'make' again, the build directory of the
same version of the package will, by default, be deleted. Set the
environment variable SAGE_KEEP_BUILT_SPKGS=yes to prevent this.

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

The following folder contains config.log and install.log files

Thank you

--Nasser

[François Bissey]

There are at least two different issues that breaks the build of the doc
for you:

[sagemath_doc_html-none] pkg_resources.DistributionNotFound: The
'fpylll<=0.5.9,>=0.5.9' distribution was not found and is required by
sagemath-standard
[sagemath_doc_html-none] Warning: Could not import
sage.manifolds.differentiable.characteristic_cohomology_class
/home/me/TMP/sage-9.8.rc0/local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/scipy/special/_ufuncs.cpython-310-x86_64-linux-gnu.so:
undefined symbol: dstevr_

It seems that fpylll is not installed (yet?) and that's a bit of an issue.
The second problem is a bit more interesting 'dsterv_` is a fortran
symbol. It looks like your scipy is missing the lapack package it has
been compiled against - usually openblas - or that lapack is incomplete
for some reason or not found in the runtime library path.

François

> The following folder <https://12000.org/tmp/sagemath_logs/> contains

> https://groups.google.com/d/msgid/sage-release/d6aecaee-279a-45f8-a645-36483d72a382n%40googlegroups.com <https://groups.google.com/d/msgid/sage-release/d6aecaee-279a-45f8-a645-36483d72a382n%40googlegroups.com>.

>
> --
> 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

> <mailto:sage-release...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/sage-release/4a516a44-b761-4f02-ac10-e121cadc54d3n%40googlegroups.com <https://groups.google.com/d/msgid/sage-release/4a516a44-b761-4f02-ac10-e121cadc54d3n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Dima Pasechnik]

Hello,

Was it  an incremental build. If so, 

   make distclean

and

build again from scratch.

Anyhow, if you look in your istall.log you'll see problems with a number of packages, e.g. fpylll, scipy.

[sagemath_doc_html-none] [manifolds] ImportError: /home/me/TMP/sage-9.8.rc0/local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/scipy/special/_ufuncs.cpython-310-x86_64-linux-gnu.so: undefined symbol: dstevr_

Probably the problem is linking with openblas. Do you have a number of openblas versions accessible to the linker?

By the way,

what are you trying to achieve with exporting

MAXIMA ?

Sage does not support using external Maxima at the moment,

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/4a516a44-b761-4f02-ac10-e121cadc54d3n%40googlegroups.com.

[Nasser M. Abbasi]

"Was it  an incremental build. If so, 

   make distclean

and

build again from scratch."

It was not an incremental build. I extracted the gzip file and did the steps I showed. So I do not need to do "make distclean" since everything is new.

But I repeated now everything again using "make distclean" and got same error.

"Probably the problem is linking with openblas. Do you have a number of openblas versions accessible to the linker?"

I have no idea.  What command should I use to check? 

I am just following the instruction to download and install sagemath. 

I have giac and maxima and fricas and texlive already installed on this virtual box already. I also have sagemath 9.7 installed in different folder.

--Nasser

[Dima Pasechnik]

On Sat, 4 Feb 2023, 13:51 'Nasser M. Abbasi' via sage-release, <sage-r...@googlegroups.com> wrote:

"Was it  an incremental build. If so, 

   make distclean

and

build again from scratch."

It was not an incremental build. I extracted the gzip file and did the steps I showed. So I do not need to do "make distclean" since everything is new.

But I repeated now everything again using "make distclean" and got same error.

"Probably the problem is linking with openblas. Do you have a number of openblas versions accessible to the linker?"

I have no idea.  What command should I use to check? 

I am just following the instruction to download and install sagemath. 

I have giac and maxima and fricas and texlive already installed on this virtual box already. I also have sagemath 9.7 installed in different folder.

sage in a different directory which is in your PATH typically breaks things.

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/4d86cb2d-8c03-4101-904e-21a1a1ce8b93n%40googlegroups.com.

[Dima Pasechnik]

On Sat, Feb 4, 2023 at 1:51 PM 'Nasser M. Abbasi' via sage-release
<sage-r...@googlegroups.com> wrote:
>
> "Was it an incremental build. If so,
> make distclean
> and
> build again from scratch."
>
> It was not an incremental build. I extracted the gzip file and did the steps I showed. So I do not need to do "make distclean" since everything is new.
>
> But I repeated now everything again using "make distclean" and got same error.
>
> "Probably the problem is linking with openblas. Do you have a number of openblas versions accessible to the linker?"
>
> I have no idea. What command should I use to check?

run

ldd /home/me/TMP/sage-9.8.rc0/local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/scipy/special/_ufuncs.cpython-310-x86_64-linux-gnu.so

Post its output
- there should be an openblas mentioned, the question is which one.

As far as fpylll is concerned, you've hit the bug fixed in not yet merged
https://trac.sagemath.org/ticket/34940

You should add fpylll at the end of the line in

build/pkgs/sagemath_doc_html/dependencies

and run ./configure and make again

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/4d86cb2d-8c03-4101-904e-21a1a1ce8b93n%40googlegroups.com.

[Dima Pasechnik]

one source of trouble might be your

LD_LIBRARY_PATH=/usr/local/lib:/mnt/g/public_html/scripts:

setting. Build systems are usually ignoring it, linking with "correct"
libraries,
but at runtime (e.g., while importing a Python module using libraries)
this might result in picking a wrong library.

[Nasser M. Abbasi]

Here is the output

>ldd /home/me/TMP/sage-9.8.rc0/local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/scipy/special/_ufuncs.cpython-310-x86_64-linux-gnu.so
    linux-vdso.so.1 (0x00007ffc53be9000)
    libgfortran.so.5 => /usr/lib/libgfortran.so.5 (0x00007fc232878000)
    libm.so.6 => /usr/lib/libm.so.6 (0x00007fc232790000)
    libmvec.so.1 => /usr/lib/libmvec.so.1 (0x00007fc232696000)
    libgcc_s.so.1 => /usr/lib/libgcc_s.so.1 (0x00007fc232676000)
    libc.so.6 => /usr/lib/libc.so.6 (0x00007fc23248f000)
    libquadmath.so.0 => /usr/lib/../lib/libquadmath.so.0 (0x00007fc232444000)
    /usr/lib64/ld-linux-x86-64.so.2 (0x00007fc232ce1000)
>

"You should add fpylll at the end of the line in

build/pkgs/sagemath_doc_html/dependencies

and run ./configure and make again"

Ok thanks. Will give this a try.

"sage in a different directory which is in your PATH typically breaks things"

but it is NOT in the path. This is the reason I do 

        unset SAGE_ROOT 

       unset SAGE_LOCAL 

before building any new release. I make sure current installed sagemath (in different folder) is not visible.

--Nasser

[Dima Pasechnik]

On Sun, Feb 5, 2023 at 9:04 AM 'Nasser M. Abbasi' via sage-release
<sage-r...@googlegroups.com> wrote:
>
>
> Here is the output
>
> >ldd /home/me/TMP/sage-9.8.rc0/local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/scipy/special/_ufuncs.cpython-310-x86_64-linux-gnu.so
> linux-vdso.so.1 (0x00007ffc53be9000)
> libgfortran.so.5 => /usr/lib/libgfortran.so.5 (0x00007fc232878000)
> libm.so.6 => /usr/lib/libm.so.6 (0x00007fc232790000)
> libmvec.so.1 => /usr/lib/libmvec.so.1 (0x00007fc232696000)
> libgcc_s.so.1 => /usr/lib/libgcc_s.so.1 (0x00007fc232676000)
> libc.so.6 => /usr/lib/libc.so.6 (0x00007fc23248f000)
> libquadmath.so.0 => /usr/lib/../lib/libquadmath.so.0 (0x00007fc232444000)
> /usr/lib64/ld-linux-x86-64.so.2 (0x00007fc232ce1000)

here is how it looks on a non-broken installation.

$ ldd local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/scipy/special/_ufuncs.cpython-310-x86_64-linux-gnu.so
linux-vdso.so.1 (0x00007ffe6b935000)
libopenblas.so.0 => /lib64/libopenblas.so.0 (0x00007fe561400000)
libgfortran.so.5 => /lib64/libgfortran.so.5 (0x00007fe561000000)
libm.so.6 => /lib64/libm.so.6 (0x00007fe563865000)
libmvec.so.1 => /lib64/libmvec.so.1 (0x00007fe561303000)
libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007fe563845000)
libc.so.6 => /lib64/libc.so.6 (0x00007fe560c00000)
libquadmath.so.0 => /lib64/libquadmath.so.0 (0x00007fe560fb8000)
/lib64/ld-linux-x86-64.so.2 (0x00007fe563ad3000)

As you see, in your case libopenblas.so is not there, and you have an
error due to this.
Why scipy builds such a broken module, I don't know. I'd say it's a scipy bug.

Please post logs/pkgs/numpy* and logs/pkgs/scipy*

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/2660c6b9-57cd-4d75-b0f8-97636bbfaf71n%40googlegroups.com.

[Dima Pasechnik]

please also check that

ldd local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/numpy/core/_multiarray_umath.cpython-310-x86_64-linux-gnu.so

has openblas in its output.

[Nasser M. Abbasi]

Here is the output for the new command you posted

>ldd local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/numpy/core/_multiarray_umath.cpython-310-x86_64-linux-gnu.so
    linux-vdso.so.1 (0x00007ffd883d4000)
    libcblas.so.3 => /usr/lib/libcblas.so.3 (0x00007f25a6532000)
    libopenblas.so.3 => /usr/lib/libopenblas.so.3 (0x00007f25a5260000)
    libm.so.6 => /usr/lib/libm.so.6 (0x00007f25a5178000)
    libc.so.6 => /usr/lib/libc.so.6 (0x00007f25a4f91000)
    /usr/lib64/ld-linux-x86-64.so.2 (0x00007f25a6a5b000)
    libgomp.so.1 => /usr/lib/libgomp.so.1 (0x00007f25a4f48000)
    libgcc_s.so.1 => /usr/lib/libgcc_s.so.1 (0x00007f25a4f28000)
>

I've also put in same folder here  a file called log_folders.tar.gz  which is backup of the logs/ folder in my build tree of 9.8 rc0 so it will have the logs you asked for.

any other information needed please let me know.

--Nasser

[Dima Pasechnik]

I suspect you have an installation of openblas, or/and some other
blas/lapack, in /usr/local - and this results in scipy being unable
to link it correctly.


(or perhaps it's not /usr/local, but some other location known to pkg-config)

numpy and scipy accumulated ~20 years of layer upon layer of old
outdated code for recognising BLAS/LAPACK libraries, it's such a
nightmare to debug.



On Sun, Feb 5, 2023 at 10:36 AM 'Nasser M. Abbasi' via sage-release

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/f362b0db-0f36-4c03-9949-473954352652n%40googlegroups.com.

[Nasser M. Abbasi]

I agree with you. I think the best thing for me, is just to make a brand new Virtual box, and install a new Linux and start everything from clean system. It will take me 2-3 days as I have to install few things I need, Then will try installing sagemath again after that.

Thanks for the help.

--Nasser

[Dima Pasechnik]

On Sun, Feb 5, 2023 at 12:44 PM 'Nasser M. Abbasi' via sage-release
<sage-r...@googlegroups.com> wrote:
>
>
> I agree with you. I think the best thing for me, is just to make a brand new Virtual box, and install a new Linux and start everything from clean system. It will take me 2-3 days as I have to install few things I need, Then will try installing sagemath again after that.

you can try to temporarily move /usr/local somewhere, and see if it
helps; no need to rebuild the whole Sage then, just rebuild
scipy:

make scipy-clean scipy-uninstall
make scipy

Then start sage

./sage

and run

sage: import scipy

there. If this works, run

make

to see if it finishes fine.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/dcab8085-34af-40db-aaf7-e26db1c4dae6n%40googlegroups.com.

[Dima Pasechnik]

You also were saying that you have MAXIMA set to an installation of Maxima.

Note that Sage won't be able to use it. Sage uses Maxima in two ways:
as a Python module embedding ECL (with Maxima installed there)
and (something we work on to get rid of) interacting with Maxima built
with ECL with pexpect (Unix pipes, more or less).
While the latter may in principle use different Maxima installation,
this is not implemented.