Sage 9.1.rc3 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 

There are no more open blocker tickets no now would be a good time for you to test if you haven't already!

bfbaddcc58 (tag: 9.1.rc3, trac/develop) Updated SageMath version to 9.1.rc3

ec4fd4a567 Trac #29625: support for weighted term orders in normal_basis

6e2957b534 Trac #29348: Remove dependency on 'twisted' for some unpickling in cartan_type

d51fbf9df3 Trac #28753: Ramification Type for Rational Maps

fa25d32e2e Trac #24601: 3D plot missing axis labels and tick numbers with viewer='threejs' and displayed in chrome.

f0f8016277 Trac #29624: PermutationGroup_generic.__richcmp__ does not check for op_NE

c8c581e29b Trac #29609: Compute the Conway mass directly from the genus symbol.

dc102aa724 Trac #29604: Laurent polynomials rings in characteristic 2 have strange generators

6bf0780980 Trac #29592: change _cmp_ to _richcmp_ in finite rings

013ce41106 Trac #29570: Wrong parent when using diff_form of degree zero

951a41f7b3 Trac #29553: add a method to test whether a polynomial is symmetric

1b358237ec Trac #25797: Implement graftal lace cellular automata

17e126688a Trac #29627: "sage -python", "sage -pip", etc.: Do not fall back to system Python

a4a63c8349 Trac #29626: Remove ./configure --enable-SPKG options for pip packages (for which this is not implemented)

e4fd7dfe8a Trac #29618: build/pkgs/conda.txt: Use python<3.8

1924002fd7 Trac #29612: Fix palp issue with -DNDEBUG

ec976ffd37 Trac #29606: build/pkgs/maxima/spkg-install.in: Do not pass on another copy of CFLAGS, LDFLAGS

d92d65d89f Trac #29601: Fix singular configure patch

4256058c21 Trac #29599: Remove C++17 flag from giac and lcalc

20f930cead Trac #29598: Make sure latest openblas from conda is used

588010006e Trac #29597: Don't override LDFLAGS in m4rie

84a3744d65 Trac #29596: Add rpath-link on Linux

4f52bf1137 Trac #29621: remove log_html() and other log_*()

903edac714 Trac #29607: homebrew-standard: Fix several problems with library search paths

cc17d2e22b Trac #29448: configure finds libpng but matplotlib does not

1302e46472 Trac #29269: Add call to cliquer with custom cliques sizes

31666f8b04 Trac #29603: wrong self import in skew polynomials

990993d2cc Trac #29593: tuning our .lgtm.yml

425d099420 Trac #29579: internet test in OEIS failing

dd2c3bace3 Trac #29578: failing internet test integration

876a534f69 Trac #29594: Remove some dead spkg-configure.m4 code

ee9518ff3e Trac #29591: Don't test whether ppl is an installed package

d65f415d62 Trac #29587: build/pkgs/glpk/distros/homebrew.txt: Enable glpk again

ba00d2a5a0 Trac #29563: perl_term_readline_gnu spkg-configure depends on TERM environment variable

1c8196af72 Trac #29561: Don't use imports from rings.all in matrices

c4d6a67946 Trac #29530: Make Docker images generated by GitHub CI workflows available; minor CI fixes

a4c7371d6a Trac #29390: gf2x build fails on fedora-32-minimal

cbcab1ad33 Trac #29176: Bug in Voronoi Diagram

f6d1907e33 Trac #26513: bug in discovery of SRG' complements - related to twograph_descendant

f4d2829903 Trac #26349: Memory leaks with matrix operations over GF(2)

e080f18c9b Trac #26287: Upgrade to gc 8.0.4 and libatomic 7.6.10

97c47d1e61 Trac #29582: polymake SPKG.txt: Remove outdated info regarding macOS headers

c05af490cd Trac #29560: sage-guess-package-system: Ignore conda if no environment is activated

790f4237bc Trac #27699: configure: Exit with error if an Anaconda environment without compilers is active

be00014c2b Trac #29584: pillow: Add dependencies

cdc2ccf376 Trac #29574: conda-forge-macos-standard: sage.libs.pynac.pynac Symbol not found

f927058305 Trac #29564: fix some details suggested by lgtm in number fields and quaternions

bb5a3ff819 Trac #29571: build/pkgs/*/distros: Add some missing system packages

89e63ff18b Trac #29567: Fix "invalid escape sequence" when building the doc

715458fd4a Trac #29466: Add kwds to __classcall__ of MatrixSpace

f1683767f9 Trac #29562: .homebrew-build-env: Set CPATH, LIBRARY_PATH instead of CPPFLAGS, LDFLAGS

b5dc372023 Trac #29556: Update README and Installation Guide (from Source) for Sage 9.1

50c5ac80b0 (tag: 9.1.rc2) Updated SageMath version to 9.1.rc2

[Samuel Lelièvre]

# Building SageMath 9.1.rc3 on Debian 9

Based on a working Sage 9.1.rc2.

## Configure

```
$ git pull origin develop --tags
[...]
$ make configure
[...]
$ ./configure
[...]
```

The `./configure` step recommends:

```
$ sudo apt-get update
$ sudo apt-get install libboost-dev coinor-cbc coinor-libcbc-dev
pari-gp2c libisl-dev libxml-libxslt-perl libxml-writer-perl
libxml2-dev libperl-dev libfile-slurp-perl libjson-perl libsvg-perl
libterm-readkey-perl libterm-readline-gnu-perl libmongodb-perl
```

But some of them are already there:

```
$ dpkg-query --list libboost-dev coinor-cbc coinor-libcbc-dev
pari-gp2c libisl-dev libxml-libxslt-perl libxml-writer-perl
libxml2-dev libperl-dev libfile-slurp-perl libjson-perl libsvg-perl
libterm-readkey-perl libterm-readline-gnu-perl libmongodb-perl
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name Version Architecture Description
+++-========================-=================-=================-=====================================================
ii coinor-cbc 2.8.12-1+b2 amd64
Coin-or branch-and-cut mixed integer programming solv
ii coinor-libcbc-dev 2.8.12-1+b2 amd64
Coin-or branch-and-cut mixed integer programming solv
ii libboost-dev:amd64 1.62.0.1 amd64 Boost
C++ Libraries development files (default versio
ii libisl-dev:amd64 0.18-1 amd64
manipulating sets and relations of integer points bou
ii libperl-dev 5.24.1-3+deb9u6 amd64 Perl
library: development files
ii libterm-readline-gnu-per 1.35-1 amd64 Perl
extension for the GNU ReadLine/History Library
ii libxml-libxslt-perl 1.95-1+b1 amd64 Perl
interface to the GNOME libxslt library
ii libxml-writer-perl 0.625-1 all Perl
module for writing XML documents
ii libxml2-dev:amd64 2.9.4+dfsg1-2.2+d amd64
Development files for the GNOME XML library
ii pari-gp2c 0.0.10-1 amd64
PARI/GP GP to C compiler
dpkg-query: no packages found matching libfile-slurp-perl
dpkg-query: no packages found matching libjson-perl
dpkg-query: no packages found matching libsvg-perl
dpkg-query: no packages found matching libterm-readkey-perl
dpkg-query: no packages found matching libmongodb-perl
```

This might be (?) in the same vein as the problem reported at

- Sage Trac ticket 29586
Improve configure recommendations
https://trac.sagemath.org/ticket/29586

## Build

```
$ make -s V=0
[...]
```

All fine.

## Tests

I wonder if it's normal that `make ptestlong` seems to
start by `make sagelib` and `make dochtml` even though
those were run just before. I forget if there's a ticket for that.

Two files with failing doctests:

```
$ make ptestlong
[...]
----------------------------------------------------------------------
sage -t --long src/sage/tests/cmdline.py # 3 doctests failed
sage -t --long src/sage/interfaces/psage.py # 1 doctest failed
----------------------------------------------------------------------
Total time for all tests: 1669.1 seconds
cpu time: 11902.3 seconds
cumulative wall time: 12716.7 seconds
Makefile:182: recipe for target 'ptestlong' failed
make: *** [ptestlong] Error 1
$
```

## Retest the files with failing doctests,

One of them(`src/sage/tests/cmdline.py`) still fails 3 doctests:

```
$ ./sage -t --long src/sage/tests/cmdline.py
too many failed tests, not using stored timings
Running doctests with ID 2020-05-06-00-31-02-ac8fcb17.
Git branch: develop
Using --optional=build,dochtml,memlimit,sage
Doctesting 1 file.
sage -t --long src/sage/tests/cmdline.py
**********************************************************************
File "src/sage/tests/cmdline.py", line 598, in
sage.tests.cmdline.test_executable
Failed example:
out.startswith("3.")
Expected:
True
Got:
False
**********************************************************************
File "src/sage/tests/cmdline.py", line 600, in
sage.tests.cmdline.test_executable
Failed example:
err
Expected:
''
Got:
'/home/lelievre/s/sage3/src/bin/sage: line 546: exec: sqlite3: not found\n'
**********************************************************************
File "src/sage/tests/cmdline.py", line 602, in
sage.tests.cmdline.test_executable
Failed example:
ret
Expected:
0
Got:
127
**********************************************************************
1 item had failures:
3 of 233 in sage.tests.cmdline.test_executable
[232 tests, 3 failures, 53.14 s]
----------------------------------------------------------------------
sage -t --long src/sage/tests/cmdline.py # 3 doctests failed
----------------------------------------------------------------------
Total time for all tests: 53.2 seconds
cpu time: 0.6 seconds
cumulative wall time: 53.1 seconds
$
```

The other one (`src/sage/interfaces/psage.py`) passes all tests:

```
$ ./sage -t --long src/sage/interfaces/psage.py
too many failed tests, not using stored timings
Running doctests with ID 2020-05-06-00-33-48-cfa1037b.
Git branch: develop
Using --optional=build,dochtml,memlimit,sage
Doctesting 1 file.
sage -t --long src/sage/interfaces/psage.py
[6 tests, 7.02 s]
----------------------------------------------------------------------
All tests passed!
----------------------------------------------------------------------
Total time for all tests: 7.1 seconds
cpu time: 0.1 seconds
cumulative wall time: 7.0 seconds
```

[Emmanuel Charpentier]

On Debian testing running on core i7 + 16 GB RAM, rebuilding 9.1.rc2 afresh (make distclean, make config , ./configure, make ptestlong) leads to ten permanent errors:

File	Result	P/T
src/sage/schemes/elliptic_curves/ell_number_field.py	3 doctests failed	P
src/sage/rings/number_field/number_field.py	7 doctests failed	P
src/sage/rings/number_field/number_field_element.pyx	1 doctest failed	P
src/sage/rings/number_field/number_field_ideal.py	2 doctests failed	P
src/sage/lfunctions/dokchitser.py	2 doctests failed	P
src/sage/numerical/backends/glpk_backend.pyx	1 doctest failed	P
src/sage/rings/polynomial/polynomial_quotient_ring.py	2 doctests failed	P
src/sage/lfunctions/pari.py	1 doctest failed	P
src/sage/libs/glpk/error.pyx	1 doctest failed	P
src/sage/rings/number_field/unit_group.py	1 doctest failed	P

The results of re-running these tests is in the attached file chkerrs.txt.

HTH,

[Emmanuel Charpentier]

Of note : ./configure suggested to install libcdd-dev and libcdd-tools, which were already installed :

charpent@zen-book-flip:/usr/local/sage-9$ sudo apt-get install libcdd-dev libcdd-tools

Lecture des listes de paquets... Fait

Construction de l'arbre des dépendances       

Lecture des informations d'état... Fait

libcdd-dev est déjà la version la plus récente (094j-2).

libcdd-tools est déjà la version la plus récente (094j-2).

0 mis à jour, 0 nouvellement installés, 0 à enlever et 0 non mis à jour.

HTH,

[Matthias Köppe]

On Wednesday, May 6, 2020 at 10:27:03 AM UTC-7, Emmanuel Charpentier wrote:

Of note : ./configure suggested to install libcdd-dev and libcdd-tools, which were already installed

I've just added the following to README.md in https://trac.sagemath.org/ticket/25206:

   Sometimes, the messages will recommend to install packages that are

   already installed on your system. See the earlier configure

   messages or the file `config.log` for explanation.  Also, the

   messages may recommend to install packages that are actually not

   available; only the most recent releases of your distribution will

   have all of these recommended packages.

 

[Jonathan Kliem]

Thank you Samuel Lelievre and Emmanuel Charpentier for testing. Your doctests failures are known:

If pari 2.11.4 is installed on the system (e.g. debian testing), there are some failing doctests due to different representatives (see #29472 and #29313). This is really nothing to worry about just annoying:

sage -t src/sage/lfunctions/dokchitser.py # 2 doctests failed
sage -t src/sage/lfunctions/pari.py # 1 doctest failed
sage -t src/sage/rings/polynomial/polynomial_quotient_ring.py # 2 doctests failed
sage -t src/sage/schemes/elliptic_curves/ell_number_field.py # 3 doctests failed
sage -t src/sage/rings/number_field/unit_group.py # 1 doctest failed
sage -t src/sage/rings/number_field/number_field_element.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

If you use glpk from your system, you will have two failing doctests as upstream rejected a patch that improves the error traceback for sage/giving warnings (to my understanding, see #29493):

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

If you have libsqlite3 but not sqlite3 installed, you don't have the sqlite3 executable, witnessed by the following test (#29092):

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

I think none of those failures is serious enough that it needs to be handled now. We also need to make a general decision on how we deal with doctests that may vary slighly by the system package and how rigid we are with accepting system packages (is the glpk "bug" already serious enough to reject it).

Jonathan

[Sébastien Labbé]

Testing few optional and external packages, I get:

Using --optional=4ti2,cbc,ccache,cryptominisat,dot2tex,e_antic,external,fricas,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/geometry/polyhedron/base.py  # Bad exit: 1
sage -t --long src/sage/graphs/generators/smallgraphs.py  # 2 doctests failed
----------------------------------------------------------------------
External software detected for doctesting: cplex,ffmpeg,graphviz,imagemagick,internet,latex,pandoc

Rerunning failed doctests, I get the same:

----------------------------------------------------------------------
sage -t --long src/sage/databases/findstat.py  # 8 doctests 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
----------------------------------------------------------------------
External software detected for doctesting: internet

Follow up at https://trac.sagemath.org/ticket/25536

[Justin C. Walker]

> On May 5, 2020, at 15:24 , 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 fresh clones of the develop tree, on three macOS systems: 10.13.6 (2017 iMac Pro, 18-core Xeon W), 10.14.6 (2017 MBP, Quad-core Core i7), 10.15.4 (2019 MBP, 9-core Core i9).

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

Justin

--
Justin C. Walker, Curmudgeon at Large
Director
Institute for the Enhancement of the Director's income
-----------
Question 43:
What if the hokey pokey
really *is* what it’s all about?
--

[Samuel Lelievre]

Wed 2020-05-06 19:35:12 UTC, Jonathan Kliem:

>

> Thank you Samuel Lelievre and Emmanuel Charpentier

> for testing. Your doctests failures are known:

>

> If pari 2.11.4 is installed on the system

<snip>

> If you use glpk from your system,

<snip>

> If you have libsqlite3 but not sqlite3 installed,

<snip>

> I think none of those failures is serious enough that

> it needs to be handled now. We also need to make a

> general decision on how we deal with doctests that

> may vary slighly by the system package and how rigid

> we are with accepting system packages (is the glpk

> "bug" already serious enough to reject it).

Thanks Jonathan for the explanation of these points.

That helped. Got sqlite3 installed on the Debian 9

system and that got tests to pass for Sage 9.1.rc4.