Sage 10.2.beta5 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

1cf0c13e527 (tag: 10.2.beta5, github/develop) Updated SageMath version to 10.2.beta5
b01856309bc gh-36345: `ecm`: Work around build failure with Xcode 15
2c714728fd0 gh-36339: fix expect interface for newer ptyprocess
bb04839a5d5 gh-36338: CI: Merge open blocker PRs in all CI workflows + other improvements
a9cb852973a gh-36336: add some class roles for linking Errors in doc
247ea886f0e gh-36335: using more itertools.product
f4dc2bc1c8e gh-36330: src/sage/repl/load.py: replace tmp_dir()
f6a9058bca4 gh-36327: fix the linter once more
496fb917960 gh-36326: more fixes for E228 and E225 in combinat and some other folders
ccf337e8443 gh-36323: src/sage/combinat/words/words.py: replace tmp_dir()
feed3d41ef9 gh-36319: build/pkgs/setuptools_scm_git_archive: Remove (obsolete)
22a32e79929 gh-36316: build/pkgs/lrcalc_python: add standard python spkg-configure.m4
034f1f643b7 gh-36299: Generic implementation of fitting ideal
c325311ba47 gh-36284: improve checks
0b5fc8e97a0 gh-36276: Yet more spkg_configure for standard python packages
ed478e76787 gh-36275: `sage.graphs.generic_graph`: some care with return ... else statements
602b57114df gh-36270: full pep8 for modular/hecke
4f8a8af331f gh-36267: Change `ipympl`/`pkgconfig`/`widgetsnbextension` to wheel packages, drop build deps
1af7f8f2b6b gh-36234: Use patchelf from the system
ce855de61cb gh-36166: Additions to the bigraded Betti number methods
05d7d1dceeb gh-36110: Update Cython to 3.0.2
5cef467b357 gh-35866: CI build.yml, doc-build.yml: Use output groups
2829cf2182f gh-35537: Fix Sphinx markup in some file
4d3e807ba54 (tag: 10.2.beta4) Updated SageMath version to 10.2.beta4

[Emmanuel Charpentier]

On Debian testing running oncore i7 + 16 GB RAM, upgrading 10.2.beta4 to 10.2.beta5 fails ; the problems seem to be bound to Cython.

Log follows in a followup mail (can’t be atached in a Google Groups mail…).

—
Emmanuel Charpentier

​

[emanuel.c...@gmail.com]

Installation log.

HTH,

Le jeudi 28 septembre 2023 à 00:10 -0700, Emmanuel Charpentier a écrit :

---------- Message transféré ---------
De : Emmanuel Charpentier <Inconnu>
Date : jeudi 28 septembre 2023 à 09:09:26 UTC+2
Objet : Re: Sage 10.2.beta5 released
À : sage-release <Inconnu>

[Thierry Dumont]

For me, with Debian testing up to date, upgrading from beta3 to beta5
was ok, after a "make distclean".

Le 28/09/2023 à 09:09, Emmanuel Charpentier a écrit :
> On Debian testing running oncore i7 + 16 GB RAM, upgrading 10.2.beta4 to

> 10.2.beta5 /fails/ ; the problems seem to be bound to Cython.

> --
> 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/f28b9181-3a8e-4ae2-94be-334e833f3373n%40googlegroups.com <https://groups.google.com/d/msgid/sage-release/f28b9181-3a8e-4ae2-94be-334e833f3373n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Kenji Iohara]

On my Mac OS 13.6 with Intel Core i5, with updated homebrew and Xcode 15, Python 3.11.5, 

The first built has failed with the package  scipy-1.11.2 and not with ecm…. !

Here is its log file: 

[John H Palmieri]

Kenji: for me with updated homebrew etc., I can only get scipy to build if I configure Sage with `./configure --with-system-gfortran=no`.

[John H Palmieri]

As with the past few releases — and I think this is due to OS X and command-line tool upgrades, not Sage — I have been getting a number of failures due to some warning messages. These continue with 10.2.beta5, and they are addressed by https://github.com/sagemath/sage/pull/36337. The giac-related failures, perhaps starting with OS X 13.3 (https://github.com/sagemath/sage/issues/35646), are also still there.

[Emmanuel Charpentier]

As suggested by tdumont, mèmake discleanallows to rebuild 10.2.beta5. Runningptestlong` afterwards gives the same two failures reported since 10.1.rc0.

HTH,

​

[Kazuyoshi Furutaka]

Hi...

The only thing I can do is to test the build (and use) of the software, and therefore I do so...

(useless?)

The build of the release on Fedora 38 (x86_64).

1. With system packages, the build finished w/o any problem.

2. The build by following the procedure described in

https://doc.sagemath.org/html/en/installation/conda.html#using-conda-to-provide-system-packages-for-the-sage-distribution

(with environment-optional.yml in place of environment.yml) stopped with the following errors while doing `make doc-clean doc-uninstall && make -j10`:

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

* package:         pythran-0.14.0
  last build time: Sep 28 19:36
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys/logs/pkgs/p
ythran-0.14.0.log
  build directory: /home/furutaka/mambaforge/envs/sage-build/var/tmp/sage/build/
pythran-0.14.0

* package:         fastjsonschema-2.18.0
  last build time: Sep 28 19:36
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys/logs/pkgs/fastjsonschema-2.18.0.log
  build directory: /home/furutaka/mambaforge/envs/sage-build/var/tmp/sage/build/fastjsonschema-2.18.0

* package:         sage_conf-10.2.beta5
  last build time: Sep 28 19:36
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys/logs/pkgs/sage_conf-10.2.beta5.log

The former two seemed to be caused because of the lack of setuptools???

Kazuyoshi

2023年9月28日(木) 7:56 Volker Braun <vbrau...@gmail.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.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/b58048a8-9158-4c56-ac41-5fbd403938a9n%40googlegroups.com.

--

Kazuyoshi Furutaka

[Matthias Köppe]

Thanks for sharing the logs for these installation problems; this is valuable information. 

I agree that there seems to be something wrong with the installation of setuptools in this environment. The command "conda list" could provide additional helpful information.

You could try if "make setuptools_wheel-no-deps" improves the situation.

There's also an open PR https://github.com/sagemath/sage/pull/35593 with fixes for the conda environment, which may be worth trying.

[Kazuyoshi Furutaka]

Hi again...

I'll attach the output of `conda list` to this email.

And...

`make setuptools_wheel-no-deps` failed with one more error (an additional log attached):

The following package(s) may have failed to build (not necessarily

during this run of 'make setuptools_wheel-no-deps'):

* package:         pythran-0.14.0
  last build time: Sep 28 19:36
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys/logs/pkgs/pythran-0.14.0.log
  build directory: /home/furutaka/mambaforge/envs/sage-build/var/tmp/sage/build/pythran-0.14.0

* package:         fastjsonschema-2.18.0
  last build time: Sep 28 19:36
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys/logs/pkgs/fastjsonschema-2.18.0.log
  build directory: /home/furutaka/mambaforge/envs/sage-build/var/tmp/sage/build/fastjsonschema-2.18.0

* package:         sage_conf-10.2.beta5
  last build time: Sep 28 19:36
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys/logs/pkgs/sage_conf-10.2.beta5.log

* package:         setuptools_wheel-65.6.3
  last build time: Sep 29 10:51
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys/logs/pkgs/setuptools_wheel-65.6.3.log
  build directory: /home/furutaka/mambaforge/envs/sage-build/var/tmp/sage/build/setuptools_wheel-65.6.3

Later I'll read the PR...

Kazuyoshi

2023年9月29日(金) 10:36 Matthias Köppe <matthia...@gmail.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/c52f5482-09d3-4f7e-9b2d-315a5a0ceb95n%40googlegroups.com.

--

Kazuyoshi Furutaka

[Kazuyoshi Furutaka]

There's also an open PR https://github.com/sagemath/sage/pull/35593 with fixes for the conda environment, which may be worth trying.

I'm afraid I didn't do what I should do, but anyway I've tried the following:

1. `git clone -b conda_remove_broken_packages https://github.com/mkoeppe/sage.git` and copy the tree
   
2. `./bootstrap-conda`
3. `mamba env update --file environment-optional-3.11.yml --name sage-build` within sage-build
4. `./bootstrap`
5. `./configure --with-python=$CONDA_PREFIX/bin/python --prefix=$CONDA_PREFIX`
6. `conda install openssh pycodestyle esbonio` -> nbconvert-core, prompt-toolkit, and simplegeneric were installed
7. `./config.status --recheck && ./config.status`

1. `make doc-clean doc-uninstall && make -j10`

The make failed in building setuptools...

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

* package:         setuptools-68.2.2
  last build time: Sep 29 14:52
  log file:        /home/furutaka/work/sage/sage-10.2.beta5-condaSys_mkoeppe/logs/pkgs/setuptools-68.2.2.log
  build directory: /home/furutaka/mambaforge/envs/sage-build/var/tmp/sage/build/setuptools-68.2.2

Kazuyoshi

2023年9月29日(金) 11:00 Kazuyoshi Furutaka <furutaka....@gmail.com>:

--

Kazuyoshi Furutaka

[Kazuyoshi Furutaka]

Ah, sorry, I forgot to attach the log...

2023年9月29日(金) 15:33 Kazuyoshi Furutaka <furutaka....@gmail.com>:

--

Kazuyoshi Furutaka

[Kenji Iohara]

John: Thanks ! As you mentioned, it worked with `./configure --with-system-gfortran=no`.

But ptestlong has created several doc test failed….

28/09/2023 23:30、John H Palmieri <jhpalm...@gmail.com>のメール:

-- 
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/00d8cc0b-ee7d-424e-a389-1b2130216c8bn%40googlegroups.com.

[John H Palmieri]

Some of the doctest failures that I see are being dealt with at #36364; see also #36337.