Sage 9.0.beta4 released
175 views
Skip to first unread message
Volker Braun
Nov 5, 2019, 5:53:55 PM11/5/19
to sage-release
Slightly delayed announcement!
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
b90b558fc9 (tag: 9.0.beta4, trac/develop) Updated SageMath version to 9.0.beta4
b4df16c19a Trac #28126: Fix interact in ECM interface
44c642a514 Trac #28670: better category for matrix groups over finite rings
2c2790de9a Trac #28667: remove deprecated stuff in perfect matchings
07b08de6cb Trac #28182: Linear algorithm for squares in words
d649e7fe2f Trac #28668: fix the base_ring of face_split
c06a540c87 Trac #28658: Three.js: Add popup menu for accessing additional options
4243350794 Trac #28650: Fix the dimension of PolyhedronFace
e16cfda9f1 Trac #28649: py3: flush output from FLINT error message
4d72eefaf3 Trac #28429: Add the classical construction of the 120-cell
bb8c66fda2 Trac #28673: Do not create ipython symlink
e91290907b Trac #28666: dot2tex: typo in URL for graphviz
a2a2bc991e Trac #28661: macOS SageMath.app wait for wrong message, significantly delaying notebook launch
2fcba3c877 Trac #28665: speed up StandardPermutations_descents.cardinality
4748831518 Trac #28414: p_group_cohomology does not build on Python3
d0d775cd5e Trac #25390: multivariate factorization over QQbar
51e3c01e61 Trac #28663: remove deprecated stuff in padics
2fe777d784 Trac #28662: Missing edges when visualizing Polyhedra with threejs
d94f80180b Trac #28572: conversion of quotient rings to macaulay2
b960e6998d Trac #28554: Scalar Field Restrictions
ae14d8e5ae Trac #18861: Three apparently useless polyhedron methods
6e22a7289a Trac #28659: fix #28631
d59cef4040 Trac #28633: Add __copy__ method to backend normaliz of Polyhedron
3ac27639ac Trac #28628: Tensor Fields: set_restriction Behaviour
9e0e0fdb13 Trac #28591: convert chain complexes from Macaulay2 to Sage
d2838a5011 Trac #26922: Wrong f-vector for unbounded polyhedra
fdd8a95ac9 Trac #28655: Fix typos in the method 'is_combinatorially_isomorphic' of Polyhedron
e7d08d31c2 Trac #28644: unicode art fails for empty skew partition
9627de6d80 Trac #28643: Speed up incidence matrix of polyhedra
77f4e8351f Trac #28622: Patchbot and Python 3 doctest failures
9088553945 Trac #28545: some cleanup in mod 2 tests
73e16ef281 Trac #27866: Introduce graphics insets
2c7753ac60 Trac #28334: py3: polynomial_rational_flint.pyx problem
906b12b1e2 (tag: 9.0.beta3) Updated SageMath version to 9.0.beta3
Travis Scrimshaw
Nov 5, 2019, 8:37:17 PM11/5/19
to sage-release
I think p_group_cohomology-3.3 (#28414) was not updated on the mirrors. Can you check?
Thanks,
Travis
Eric Gourgoulhon
Nov 6, 2019, 4:19:10 AM11/6/19
to sage-release
On Ubuntu 18.04 running on a bi-Xeon E5-2623 (8 cores) + 16 GB RAM computer and for a python3 build:
- incremental build (from 9.0.beta3, with -j16) OK
- all tests from ptestlong passed This is the first time that Python-3 Sage passes all the tests on this computer! (thanks to the inclusion of #28334 and #28622 in this release).
Eric.
Emmanuel Charpentier
Nov 6, 2019, 7:58:50 AM11/6/19
to sage-release
FWIW, on Debian testing running on core i5 + 8 GB RAM, ptestalllong gets similar results :
- No permanent failure (yay !)
- Eleven transient failures, all already seen in previous 9.x betas :
File | Result
----------------------------------------------+-------------------
src/sage/combinat/posets/posets.py | 1 doctest failed
src/sage/plot/animate.py | 7 doctests failed
src/sage/combinat/matrices/hadamard_matrix.py | 2 doctests failed
src/sage/misc/latex.py | 1 doctest failed
src/sage/misc/persist.pyx | 2 doctests failed
src/sage/coding/databases.py | 2 doctests failed
src/sage/repl/load.py | 1 doctest failed
src/sage/graphs/graph_latex.py | 2 doctests failed
src/sage/misc/remote_file.py | 2 doctests failed
src/sage/databases/findstat.py | 72 doctests failed
src/sage/interfaces/magma_free.py | 2 doctests failed
As far as I can, tell, I never saw these failures when doing testlong. It looks more and more like a problem of concurrency specific to ptestalllong, unrelated to Python3.
Is this ticket-worthy ?
Emmanuel Charpentier
Nov 6, 2019, 10:12:58 AM11/6/19
to sage-release
Similar results on Debian testing running on core i7 + 16 GB RAM, with the following exceptions:
- src/sage/tests/gap_packages.py : permanent failure (already reported)
- src/sage/graphs/generators/smallgraphs.py : ttransient failure (not seen on the other machine, already reported for this one).
HTH,
kcrisman
Nov 7, 2019, 8:51:48 AM11/7/19
to sage-release
I admit I haven't paid enough attention to the py3 switch on sage-devel, but this still surprised me, based on the wording at https://wiki.sagemath.org/Python3-compatible%20code I suppose that when the py3 switch for default is thrown, we should make sure to have a special request for testing from the whole community to make sure we have no weird regressions on py2 until after 9.1 (or that is what the Wiki currently says).
$ ./sage -v
SageMath version 9.0.beta4, Release Date: 2019-11-02
$ ./sage -python
Python 2.7.15 (default, Nov 6 2019, 12:35:23)
[GCC 4.2.1 Compatible Apple LLVM 7.3.0 (clang-703.0.31)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>>
kcrisman
Nov 7, 2019, 9:16:58 AM11/7/19
to sage-release
Okay, I just found https://trac.sagemath.org/ticket/28660 so sorry for the noise. But I guess my comment about testers stands.
kcrisman
Nov 7, 2019, 12:30:57 PM11/7/19
to sage-release
FWIW make ptestlong (though no optional packages)
----------------------------------------------------------------------
All tests passed!
----------------------------------------------------------------------
Total time for all tests: 10172.0 seconds
cpu time: 14976.0 seconds
cumulative wall time: 19701.3 seconds
kcrisman
Nov 7, 2019, 9:02:53 PM11/7/19
to sage-release
And with Python 3 support on beta4 on Mac OS X 10.11.6, a single verifiable doctest failure on ptestlong, of which only the first one is an actual failure (the rest depend on it). I note that if I run the first test by hand, it not only works but A and B give the long name based on pid and so forth correctly.
sage -t --long src/sage/misc/sageinspect.py
**********************************************************************
File "src/sage/misc/sageinspect.py", line 2132, in sage.misc.sageinspect._sage_getsourcelines_name_with_dot
Failed example:
cython('''
class A:
def __init__(self):
"some init doc"
pass
class B:
"some class doc"
class A(A):
pass
''')
Exception raised:
Traceback (most recent call last):
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run
self.compile_and_execute(example, compiler, test.globs)
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1123, in compile_and_execute
exec(compiled, globs)
File "<doctest sage.misc.sageinspect._sage_getsourcelines_name_with_dot[3]>", line 10, in <module>
''')
File "sage/misc/lazy_import.pyx", line 353, in sage.misc.lazy_import.LazyImport.__call__ (build/cythonized/sage/misc/lazy_import.c:3686)
return self.get_object()(*args, **kwds)
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/misc/cython.py", line 599, in cython_compile
return cython_import_all(tmpfile, get_globals(), **kwds)
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/misc/cython.py", line 489, in cython_import_all
m = cython_import(filename, **kwds)
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/misc/cython.py", line 469, in cython_import
return builtins.__import__(name)
ModuleNotFoundError: No module named '_Users_...__sage_temp_GC06259_23216_tmp_lqbw2l0c_pyx_0'
**********************************************************************
File "src/sage/misc/sageinspect.py", line 2142, in sage.misc.sageinspect._sage_getsourcelines_name_with_dot
Failed example:
B.A.__name__
Exception raised:
Traceback (most recent call last):
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run
self.compile_and_execute(example, compiler, test.globs)
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1123, in compile_and_execute
exec(compiled, globs)
File "<doctest sage.misc.sageinspect._sage_getsourcelines_name_with_dot[4]>", line 1, in <module>
B.A.__name__
NameError: name 'B' is not defined
**********************************************************************
File "src/sage/misc/sageinspect.py", line 2144, in sage.misc.sageinspect._sage_getsourcelines_name_with_dot
Failed example:
B.A.__qualname__
Exception raised:
Traceback (most recent call last):
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run
self.compile_and_execute(example, compiler, test.globs)
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1123, in compile_and_execute
exec(compiled, globs)
File "<doctest sage.misc.sageinspect._sage_getsourcelines_name_with_dot[5]>", line 1, in <module>
B.A.__qualname__
NameError: name 'B' is not defined
**********************************************************************
File "src/sage/misc/sageinspect.py", line 2146, in sage.misc.sageinspect._sage_getsourcelines_name_with_dot
Failed example:
sage_getsource(B.A)
Exception raised:
Traceback (most recent call last):
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 681, in _run
self.compile_and_execute(example, compiler, test.globs)
File "/Users/.../testpy3sage/sage-9.0.beta4-py3/sage-9.0.beta4/local/lib/python3.7/site-packages/sage/doctest/forker.py", line 1123, in compile_and_execute
exec(compiled, globs)
File "<doctest sage.misc.sageinspect._sage_getsourcelines_name_with_dot[6]>", line 1, in <module>
sage_getsource(B.A)
NameError: name 'B' is not defined
**********************************************************************
1 item had failures:
4 of 8 in sage.misc.sageinspect._sage_getsourcelines_name_with_dot
[339 tests, 4 failures, 18.87 s]
kcrisman
Nov 7, 2019, 9:11:56 PM11/7/19
to sage-release
On Thursday, November 7, 2019 at 9:02:53 PM UTC-5, kcrisman wrote:
And with Python 3 support on beta4 on Mac OS X 10.11.6, a single verifiable doctest failure on ptestlong, of which only the first one is an actual failure (the rest depend on it). I note that if I run the first test by hand, it not only works but A and B give the long name based on pid and so forth correctly.
1 item had failures:4 of 8 in sage.misc.sageinspect._sage_getsourcelines_name_with_dot[339 tests, 4 failures, 18.87 s]
And
cython('''cpdef test_funct(x,y): return''')
in sage.misc.sageinspect.sage_getsourcelines seems to have similar issues, but not consistently - sometimes all tests in this file pass. Maybe some kind of race condition? Whether I use one or multiple threads does not seem to matter.
Volker Braun
Nov 8, 2019, 2:36:23 PM11/8/19
to sage-release
fixed!
Justin C. Walker
Nov 12, 2019, 2:21:04 PM11/12/19
to sage-r...@googlegroups.com
> On Nov 5, 2019, at 14:53 , Volker Braun <vbrau...@gmail.com> wrote:
>
> Slightly delayed announcement!
>
> 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
Testing (‘ptestlong’) showed no failures on either platform!
Justin
--
Justin C. Walker
Director, Institute for the Absorption of Federal Funds
--
Taxes, n.:
Of life's two certainties, the only one
for which you can get an extension.
--
kcrisman
Nov 14, 2019, 10:24:56 PM11/14/19
to sage-release
By the way, building the Mac app with beta4 on Python 3 seems to have worked very nicely!
Reply all
Reply to author
Forward
0 new messages