Sage 9.4.rc1 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 

e0cbf2c965 (tag: 9.4.rc1, trac/develop) Updated SageMath version to 9.4.rc1

952f7f9006 Trac #31565: Build still non-portable despite SAGE_FAT_BINARY=yes because of numpy

5090d3dab1 Trac #32322: Random failure in projective_ds.py

a73ad2437f Trac #32142: Implement the alternating central extension of the Onsager Lie algebra

d81cacb611 Trac #32006: Fix links in tutorial bibliography

bd073b49ba (tag: 9.4.rc0) Updated SageMath version to 9.4.rc0

[Emmanuel Charpentier]

On Debian testing running on core i5 + 8 GB RAM, upgrading 9.4.rc0 to 9.4.rc1 and running ptestlong give three permanent failures :

sage -t --long --warn-long 92.9 --random-seed=0 src/sage/doctest/test.py  # 7 doctests failed
sage -t --long --warn-long 92.9 --random-seed=0 src/sage/tests/cmdline.py  # 2 doctests failed
sage -t --long --warn-long 92.9 --random-seed=0 src/sage/groups/perm_gps/permgroup.py  # 3 doctests failed

As far as I can tell now, these failures are identical to those already reported for a couple previous releases.

HTH,

​

[Samuel Lelièvre]

2021-08-09 12:00 UTC, Emmanuel Charpentier:

>
> On Debian testing running on core i5 + 8 GB RAM,
> upgrading 9.4.rc0 to 9.4.rc1 and running ptestlong

> gives three permanent failures :

>
> sage -t --long --warn-long 92.9 --random-seed=0 src/sage/doctest/test.py # 7 doctests failed
> sage -t --long --warn-long 92.9 --random-seed=0 src/sage/tests/cmdline.py # 2 doctests failed
> sage -t --long --warn-long 92.9 --random-seed=0 src/sage/groups/perm_gps/permgroup.py # 3 doctests failed
>
> As far as I can tell now, these failures are identical to those already reported for a couple previous releases.

Regarding the last one, I opened

- Sage Trac ticket 32294
Fix failing doctests in groups/perm_gps/permgroup.py
https://trac.sagemath.org/ticket/32294

Is it the same doctests as in that ticket you see failing?
Do you have any optional packages installed?
Can you reproduce on a fresh clone?

I observed the doctest failures on Debian 10 "buster"
for Sage with many optional packages installed. That Sage
was built a few releases ago, and those tests have been
failing for a while on it, including after upgrading to 9.4.rc1.

On the same machine, another Sage, also with many optional
packages, but installed from a fresh clone more recently,
passed those doctests, both for Sage 9.4.rc0 and Sage 9.4.rc1.

I guess I should inspect the config.log files to try and spot
the key difference between these two Sage installations.

[Samuel Lelièvre]

Volker Braun:

>
> e0cbf2c965 (tag: 9.4.rc1, trac/develop) Updated SageMath version to 9.4.rc1
> 952f7f9006 Trac #31565: Build still non-portable despite SAGE_FAT_BINARY=yes because of numpy
> 5090d3dab1 Trac #32322: Random failure in projective_ds.py
> a73ad2437f Trac #32142: Implement the alternating central extension of the Onsager Lie algebra
> d81cacb611 Trac #32006: Fix links in tutorial bibliography
> bd073b49ba (tag: 9.4.rc0) Updated SageMath version to 9.4.rc0

Thanks for this new release!
I hope the last two positively reviewed
"critical" tickets can make it in Sage 9.4.

On macOS 10.14.6 "Mojave", with lots of homebrew packages,
upgrading to Sage 9.4.rc1 from a working 9.4.rc0:

```
$ source .homebrew-build-env
$ ./bootstrap -q
$ ./configure -q
configure: WARNING: ".../build/pkgs/backports_ssl_match_hostname/type"
is missing. Leftovers from another branch?
$ make -s V=0
...
Sage build/upgrade complete!
```

```
$ make -s V=0 ptestlong
...
----------------------------------------------------------------------
sage -t --long --warn-long 145.3 --random-seed=0
src/sage/geometry/polyhedron/base.py # 1 doctest failed
----------------------------------------------------------------------
Total time for all tests: 8335.9 seconds
cpu time: 25308.4 seconds
cumulative wall time: 32261.4 seconds
Pytest is not installed, skip checking tests that rely on it.
make: *** [ptestlong] Error 1

$ ./sage -t --long --warn-long 145.3 --random-seed=0
src/sage/geometry/polyhedron/base.py
Running doctests with ID 2021-08-10-04-23-28-4a95405e.
Git branch: develop
Using --optional=build,dochtml,homebrew,pip,sage,sage_spkg
Doctesting 1 file.
sage -t --long --warn-long 145.3 --random-seed=0
src/sage/geometry/polyhedron/base.py
**********************************************************************
File "src/sage/geometry/polyhedron/base.py", line 6653, in
sage.geometry.polyhedron.base.Polyhedron_base.face_lattice
Failed example:
n == get_memory_usage()
Expected:
True
Got:
False
**********************************************************************
1 item had failures:
1 of 42 in sage.geometry.polyhedron.base.Polyhedron_base.face_lattice
[1758 tests, 1 failure, 98.31 s]
----------------------------------------------------------------------
sage -t --long --warn-long 145.3 --random-seed=0
src/sage/geometry/polyhedron/base.py # 1 doctest failed
----------------------------------------------------------------------
Total time for all tests: 98.7 seconds
cpu time: 54.0 seconds
cumulative wall time: 98.3 seconds
Pytest is not installed, skip checking tests that rely on it.
```

Maybe related to
https://trac.sagemath.org/ticket/31623

[Enrique Artal]

I have the develop version installed in two computers with Fedora 34. On one of them, the update was successful, but in the second one, since it did not start I erased the installation and clone the git repository again. This is the result:

Checking whether SageMath should install SPKG gp2c...
checking whether any of pari is installed as or will be installed as SPKG... no
configure: gp prefix is /usr
configure: error: using system pari and unable to locate pari.cfg

I have no clue why this happens. Thanks, Enrique.

[Matthias Köppe]

I have opened https://trac.sagemath.org/ticket/32359 for this -- clearly a blocker for Sage 9.4

[Matthias Köppe]

Thanks, Volker, for the latest release candidate.

We still have problems on the following platforms:

- opensuse-tumbleweed (https://github.com/sagemath/sage/runs/3274952722): Errors building cypari, ecm, singular

- conda-forge: Multiple problems, as noted earlier

- fedora-34: (fixed by https://trac.sagemath.org/ticket/32359, needs review)

- centos-7-i386 (https://github.com/sagemath/sage/runs/3274952890):  SIGFPE while building dochtml

Also there remain race conditions in Python package installation - fixed in https://trac.sagemath.org/ticket/32361 (needs review)

As we are preparing notes on platform support for https://wiki.sagemath.org/ReleaseTours/sage-9.4, it would be good to know about the status of Apple Silicon (M1) support (https://trac.sagemath.org/ticket/30592).

[Volker Braun]

I also get the error in get_memory_usage, mostly on OSX. Definitely a flakey test!

[Kenji Iohara]

I could compile it from the first built on my Mac OS 11.15.1, with Intel Core i5, 

but for test or ptestlong, it failed :

[Justin C. Walker]

> On Aug 8, 2021, at 15:06 , 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 w/o problems on three macOS systems. Testing with ‘ptestlong'

10.13.6 (2017 iMac Pro, 2.3 GHz Intel Xeon W). Testing got one failure: the usual get_memory_usage() test, intermittent.

10.14.6 (2017 MBP, 3.1 GHz 4-core Core i7). Testing got one failure: external.py (log attached as 10.14.txt). Seems persistent.

10.15.7 (2019 MBP, 2.4 GHz 8-core Core i9). All tests passed!

[Matthias Köppe]

The problem on opensuse-tumbleweed is fixed by https://trac.sagemath.org/ticket/32368, needs review.

[Emmanuel Charpentier]

Sorry for the late answer...

Le lundi 9 août 2021 à 18:16:23 UTC+2, Samuel Lelievre a écrit :

2021-08-09 12:00 UTC, Emmanuel Charpentier:
>
> On Debian testing running on core i5 + 8 GB RAM,
> upgrading 9.4.rc0 to 9.4.rc1 and running ptestlong
> gives three permanent failures :
>
> sage -t --long --warn-long 92.9 --random-seed=0 src/sage/doctest/test.py # 7 doctests failed
> sage -t --long --warn-long 92.9 --random-seed=0 src/sage/tests/cmdline.py # 2 doctests failed
> sage -t --long --warn-long 92.9 --random-seed=0 src/sage/groups/perm_gps/permgroup.py # 3 doctests failed
>
> As far as I can tell now, these failures are identical to those already reported for a couple previous releases.

Regarding the last one, I opened

- Sage Trac ticket 32294
Fix failing doctests in groups/perm_gps/permgroup.py
https://trac.sagemath.org/ticket/32294

Is it the same doctests as in that ticket you see failing?

Yep. But on another system (Debian testing running on core I7 + 16 GB RAM) it passes. Go figure...

It seems to depend on atmospheric pressure and temperature, phase of the moon AND Dow Jones closing...

 

Do you have any optional packages installed?

Yep : cmdstanpy cypari2 dot2tex fricas gap_jupyter gap_package iniconfig json5 jupyter_kernel_gap jupyter_kernel_singular kenzo libsemigroups pandas pycodestyle pyflakes pysingular pytest python_dateutil pyyaml relint saclib sagemath_standard singular_jupyter tqdm ujson 

 

Can you reproduce on a fresh clone?

I'll have to install a new tree...

[Dima Pasechnik]

note that gap_package installed alters code paths of GAP code which might not even depend explicitly on 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/94a41093-7ca7-4894-8d28-bb9e540e5fb8n%40googlegroups.com.