SageMath 10.7 on macOS 15.5 (again) #40512

GMS103 · 2025-07-30T14:20:01Z

This follows #40501.

GMS103 · 2025-07-30T16:00:24Z

SageMath version 10.7.beta9 + #40512 gives the following.

macOS 15.6
After ./configure --enable-gap_packages --enable-libsemigroups --enable-symengine,
make succeeds.
And make ptestlong gives only 1 error:
sage -t --long --warn-long 30.0 --random-seed=… src/sage/algebras/quantum_groups/quantum_group_gap.py # 1 doctest failed
related to TikZ.

Debian 12
After ./configure --enable-gap_packages --enable-libsemigroups --enable-symengine,
make fails for qhull:

[spkg-install] Configuring qhull-2020-src-8.0.2 with cmake
[spkg-install] CMake Error at CMakeLists.txt:70 (cmake_minimum_required):
[spkg-install]   CMake 4.0 or higher is required.  You are running version 3.25.1
[spkg-install] 
[spkg-install] 
[spkg-install] -- Configuring incomplete, errors occurred!
[spkg-install] ********************************************************************************
[spkg-install] Error configuring qhull-2020-src-8.0.2 with cmake
[spkg-install] ********************************************************************************
************************************************************************
Error installing package qhull-2020-src-8.0.2
************************************************************************

After replacing minimum required cmake version 4.0 by 3.25,
make succeeds.
And make ptestlong succeeds too.
There is cmake 4.0.3-1 (experimental), I have not tried it.

dimpase · 2025-07-30T20:29:39Z

On Gentoo Linux I have a problem building guava (one of GAP packages in gap_packages). gcc-15 needs a newer version than the one in gap 4.14.0. (3.20 or newer is needed). I get

[spkg-install] cd ./src/leon; autoreconf --install --force ; ./configure; make CC="gcc" CPPFLAGS="" CFLAGS="-g -O2 -Wno-implicit-function-declaration -Wno-implicit-function-declaration -Wno-implicit-function-declaration -Wno-implicit-function-declaration -O2" LDFLAGS="-Wl,-rpath-link,/home/dima/software/sage/local/lib -L/home/dima/software/sage/local/lib -Wl,-rpath,/home/dima/software/sage/local/lib -Wl,-rpath-link,/home/dima/software/sage/local/lib -L/home/dima/software/sage/local/lib -Wl,-rpath,/home/dima/software/sage/local/lib ";
[spkg-install] In file included from src/ctjhai/minimum-weight-gf2.h:18,
[spkg-install]                  from src/ctjhai/minimum-weight.c:33:
[spkg-install] src/ctjhai/types.h:27:16: error: cannot use keyword 'false' as enumeration constant
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                ^~~~~
[spkg-install] src/ctjhai/types.h:27:16: note: 'false' is a keyword with '-std=c23' onwards
[spkg-install] src/ctjhai/types.h:27:38: error: expected ';', identifier or '(' before 'bool'
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                                      ^~~~
[spkg-install] src/ctjhai/types.h:27:38: warning: useless type name in empty declaration
[spkg-install] In file included from src/ctjhai/minimum-weight-gf2.h:18,
[spkg-install]                  from src/ctjhai/minimum-weight-gf2.c:39:
[spkg-install] src/ctjhai/types.h:27:16: error: cannot use keyword 'false' as enumeration constant
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                ^~~~~
[spkg-install] src/ctjhai/types.h:27:16: note: 'false' is a keyword with '-std=c23' onwards
[spkg-install] src/ctjhai/types.h:27:38: error: expected ';', identifier or '(' before 'bool'
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                                      ^~~~
[spkg-install] src/ctjhai/types.h:27:38: warning: useless type name in empty declaration
[spkg-install] make[3]: *** [<builtin>: src/ctjhai/minimum-weight.o] Error 1
[spkg-install] make[3]: *** Waiting for unfinished jobs....
[spkg-install] In file included from src/ctjhai/minimum-weight-gf3.h:18,
[spkg-install]                  from src/ctjhai/minimum-weight-gf3.c:57:
[spkg-install] src/ctjhai/types.h:27:16: error: cannot use keyword 'false' as enumeration constant
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                ^~~~~
[spkg-install] src/ctjhai/types.h:27:16: note: 'false' is a keyword with '-std=c23' onwards
[spkg-install] src/ctjhai/types.h:27:38: error: expected ';', identifier or '(' before 'bool'
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                                      ^~~~
[spkg-install] src/ctjhai/types.h:27:38: warning: useless type name in empty declaration
[spkg-install] make[3]: *** [<builtin>: src/ctjhai/minimum-weight-gf2.o] Error 1
[spkg-install] make[3]: *** [<builtin>: src/ctjhai/minimum-weight-gf3.o] Error 1

GMS103 · 2025-07-30T21:07:13Z

Update.

With
cmake_minimum_required(VERSION 3.25...4.0.3) (present situation)
SageMath version 10.7.beta9 + #40512 gives the following after
./configure --enable-gap_packages --enable-libsemigroups --enable-symengine

macOS 15.6
make succeeds.
make ptestlong gives only 1 error:
sage -t --long --warn-long 30.0 --random-seed=… src/sage/algebras/quantum_groups/quantum_group_gap.py # 1 doctest failed
related to TikZ (in fact there are some other failed tests, but they pass standalone).

Debian 12
make succeds.
make ptestlong succeeds too.

dimpase · 2025-07-30T22:15:48Z

you need to explicitly add --enable-semigroups !

(--enable-libsemigroups is implied then)

GMS103 · 2025-07-30T22:24:57Z

you need to explicitly add --enable-semigroups !

(--enable-libsemigroups is implied then)

Sorry, last time I tried I got
real_configure: WARNING: unrecognized options: --enable-semigroups
so I thought it was a typo.

Redoing everything again.

dimpase · 2025-07-30T23:01:59Z

you might need to run ./bootstrap

GMS103 · 2025-07-30T23:30:31Z

you might need to run ./bootstrap

I always start from scratch (except that I reuse files from /upstream because unreliable mirrors delay everything).
For me is the simplest and less time consuming.

dimpase · 2025-07-31T00:56:54Z

not sure what you mean by "from scratch", but whenever anything in build/ is changed, ./bootstrap must be successfully run to update the configure script.

inability of configure to understand the enable-semigroups option is a symptom of a stale configure script.

enriqueartal · 2025-07-31T08:58:52Z

I made a change in spkg-install.in in #40484 which allowed to build guava.

On Gentoo Linux I have a problem building guava (one of GAP packages in gap_packages). gcc-15 needs a newer version than the one in gap 4.14.0. (3.20 or newer is needed). I get

[spkg-install] cd ./src/leon; autoreconf --install --force ; ./configure; make CC="gcc" CPPFLAGS="" CFLAGS="-g -O2 -Wno-implicit-function-declaration -Wno-implicit-function-declaration -Wno-implicit-function-declaration -Wno-implicit-function-declaration -O2" LDFLAGS="-Wl,-rpath-link,/home/dima/software/sage/local/lib -L/home/dima/software/sage/local/lib -Wl,-rpath,/home/dima/software/sage/local/lib -Wl,-rpath-link,/home/dima/software/sage/local/lib -L/home/dima/software/sage/local/lib -Wl,-rpath,/home/dima/software/sage/local/lib ";
[spkg-install] In file included from src/ctjhai/minimum-weight-gf2.h:18,
[spkg-install]                  from src/ctjhai/minimum-weight.c:33:
[spkg-install] src/ctjhai/types.h:27:16: error: cannot use keyword 'false' as enumeration constant
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                ^~~~~
[spkg-install] src/ctjhai/types.h:27:16: note: 'false' is a keyword with '-std=c23' onwards
[spkg-install] src/ctjhai/types.h:27:38: error: expected ';', identifier or '(' before 'bool'
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                                      ^~~~
[spkg-install] src/ctjhai/types.h:27:38: warning: useless type name in empty declaration
[spkg-install] In file included from src/ctjhai/minimum-weight-gf2.h:18,
[spkg-install]                  from src/ctjhai/minimum-weight-gf2.c:39:
[spkg-install] src/ctjhai/types.h:27:16: error: cannot use keyword 'false' as enumeration constant
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                ^~~~~
[spkg-install] src/ctjhai/types.h:27:16: note: 'false' is a keyword with '-std=c23' onwards
[spkg-install] src/ctjhai/types.h:27:38: error: expected ';', identifier or '(' before 'bool'
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                                      ^~~~
[spkg-install] src/ctjhai/types.h:27:38: warning: useless type name in empty declaration
[spkg-install] make[3]: *** [<builtin>: src/ctjhai/minimum-weight.o] Error 1
[spkg-install] make[3]: *** Waiting for unfinished jobs....
[spkg-install] In file included from src/ctjhai/minimum-weight-gf3.h:18,
[spkg-install]                  from src/ctjhai/minimum-weight-gf3.c:57:
[spkg-install] src/ctjhai/types.h:27:16: error: cannot use keyword 'false' as enumeration constant
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                ^~~~~
[spkg-install] src/ctjhai/types.h:27:16: note: 'false' is a keyword with '-std=c23' onwards
[spkg-install] src/ctjhai/types.h:27:38: error: expected ';', identifier or '(' before 'bool'
[spkg-install]    27 | typedef enum { false = 0, true = 1 } bool;
[spkg-install]       |                                      ^~~~
[spkg-install] src/ctjhai/types.h:27:38: warning: useless type name in empty declaration
[spkg-install] make[3]: *** [<builtin>: src/ctjhai/minimum-weight-gf2.o] Error 1
[spkg-install] make[3]: *** [<builtin>: src/ctjhai/minimum-weight-gf3.o] Error 1

GMS103 · 2025-07-31T09:53:59Z

"from scratch" = rm -fr sage (especially when testing for somebody else).

GMS103 · 2025-07-31T10:02:13Z

Now using the correct configure…

With
cmake_minimum_required(VERSION 3.25...4.0.3) (present situation)
SageMath version 10.7.beta9 + #40512 gives the following after
./configure --enable-gap_packages --enable-semigroups --enable-symengine

macOS 15.6
make succeeds.
make ptestlong gives 2 errors:

(1)

sage -t --long --warn-long 30.0 --random-seed=299022206910366462203725211757600829482 src/doc/en/reference/spkg/semigroups.rst
**********************************************************************
File "src/doc/en/reference/spkg/semigroups.rst", line 15, in doc.en.reference.spkg.semigroups
Failed example:
    libgap.LoadPackage("semigroups")       # optional - semigroups
Expected:
    true
Got:
    #I  ReadPackage could not read <digraphs>/gap/doc.g
    <BLANKLINE>
    true
**********************************************************************
1 item had failures:
   1 of   3 in doc.en.reference.spkg.semigroups
    [2 tests, 1 failure, 1.32s wall]
----------------------------------------------------------------------
sage -t --long --warn-long 30.0 --random-seed=… src/doc/en/reference/spkg/semigroups.rst  # 1 doctest failed
----------------------------------------------------------------------

as indicated by DIma.
(2)

sage -t --long --warn-long 30.0 --random-seed=… src/sage/algebras/quantum_groups/quantum_group_gap.py  # 1 doctest failed

related to TikZ (in fact there are some other failed tests, but they pass standalone).

Debian 12
make succeds.
make ptestlong gives the same semigroups.rst error (in fact there are some other failed tests, but they pass standalone).

@dimpase: What do you think?

GMS103 · 2025-07-31T10:08:13Z

@dimpase
In config.log:

real_configure:126448: result: gap_packages:                   optional, SPKG version 4.14.0 will be installed
real_configure:140672: result: libsemigroups:                  no suitable system package; optional, use "./configure --enable-libsemigroups" to install SPKG version 2.7.4
real_configure:154427: result: planarity:                      no suitable system package; standard, SPKG version 4.0.0.0 will be installed
real_configure:170594: result: semigroups:                     optional, SPKG version 5.5.3 will be installed

This is incoherent for libsemigroups, do you think you could correct it?

enriqueartal · 2025-07-31T10:12:41Z

I have the same error with semigroups.srt, I thought that it may come from having system gap but with the package is the same thing. I deduce that you can build guava as it is; can you with my modification? In linux gcc-15, it is not possible. For the config.log, I got the same thing but the package was installed, maybe because of dependencies.

dimpase · 2025-07-31T18:04:38Z

@dimpase
In config.log:

real_configure:126448: result: gap_packages:                   optional, SPKG version 4.14.0 will be installed
real_configure:140672: result: libsemigroups:                  no suitable system package; optional, use "./configure --enable-libsemigroups" to install SPKG version 2.7.4
real_configure:154427: result: planarity:                      no suitable system package; standard, SPKG version 4.0.0.0 will be installed
real_configure:170594: result: semigroups:                     optional, SPKG version 5.5.3 will be installed

This is incoherent for libsemigroups, do you think you could correct it?

apart from different log formatting there seems to be no issue.

PS. libsemigroups is a dependency of semigroups, so it gets installed before the latter

dimpase · 2025-07-31T18:06:41Z

I have the same error with semigroups.srt, I thought that it may come from having system gap but with the package is the same thing. I deduce that you can build guava as it is; can you with my modification? In linux gcc-15, it is not possible. For the config.log, I got the same thing but the package was installed, maybe because of dependencies.

this is an OS-dependent race condition, it appears.

dimpase · 2025-07-31T22:00:08Z

in a chain of PRs, each should either be positively reviewed, or closed

GMS103 · 2025-07-31T22:01:48Z

Thanks, Dima.

tobiasdiez · 2025-08-01T00:42:10Z

Why is conftest.py deleted here? This seems to be a mistake.

GMS103 · 2025-08-01T00:56:29Z

Why is conftest.py deleted here? This seems to be a mistake.

Sorry, it is surely my fault, but I do not know how I did it.
Should be back now.

dimpase · 2025-08-01T00:56:51Z

the usual git commit -a mistake.

The latter stages all the known to got changed files, and confest.py is left deleted whenever make invokes ./configure (something that's an off label use of GNU toolchain).

Please restore conftest.py here and elsewhere.

GMS103 · 2025-08-01T01:03:57Z

the usual git commit -a mistake. Please restore conftest.py here and elsewhere.

You must be right.
As I said, it should be back. Could you verify?

GMS103 · 2025-08-01T01:14:25Z

Thanks Dima.

Would you give me a pointer to "the usual git commit -a mistake"?
I have much to learn yet…

vbraun · 2025-08-02T09:39:17Z

Merge conflict, please fix

enriqueartal · 2025-08-02T10:36:44Z

If I am not wrong #40519 contains all the changes of this pull request and also those of #40509. Is it better to merge just #40519? Maybe with a change of title.

vbraun · 2025-08-02T15:18:15Z

can do but #40519 has the same conflict

dimpase · 2025-08-02T19:18:06Z

@GMS103 - if you give me write access to your sage develop branch then I could fix these easily.

dimpase · 2025-08-02T19:22:29Z

anyway, a conflict is in src/doc/en/prep/Logging-On.rst - the problem there was fixed in develop branch, so this file should be just left alone.

GMS103 · 2025-08-02T21:15:00Z

@vbraun
Please do not merge this, but #40519 (I am testing it ATM).

github-actions · 2025-08-02T21:53:47Z

Documentation preview for this PR (built with commit eccbb0f; changes) is ready! 🎉
This preview will update shortly after each push to this PR.

vbraun · 2025-08-03T08:45:36Z

Closing in favor of #40519

@enriqueartal

This tries to take into account sagemath#40484. @enriqueartal: Could you test it? ### ⌛ Dependencies Built on top of sagemath#40512. URL: sagemath#40519 Reported by: Guillermo Moreno-Socías Reviewer(s): Dima Pasechnik, Enrique Manuel Artal Bartolo

GMS103 · 2025-08-03T14:26:14Z

Thanks.

@enriqueartal

This tries to take into account sagemath#40484. @enriqueartal: Could you test it? ### ⌛ Dependencies Built on top of sagemath#40512. URL: sagemath#40519 Reported by: Guillermo Moreno-Socías Reviewer(s): Dima Pasechnik, Enrique Manuel Artal Bartolo

dimpase and others added 5 commits July 29, 2025 13:23

update libsemigroups to 2.7.4

758bb59

sping GAP's semigroups package into a separate spkg

2c075bb

remove obsolete comment

66fb284

40501

8ba3074

40501

e90e6ef

github-actions bot added the s: needs review label Jul 30, 2025

GMS103 mentioned this pull request Jul 30, 2025

SageMath 10.7 on macOS 15.5 #40501

Closed

5 tasks

cmake 3.25

2c2f2bb

GMS103 mentioned this pull request Jul 30, 2025

split Semigroups into a separate spkg, bump versions #40509

Merged

5 tasks

cmake 3.25...4.0.3

746cd6b

GMS103 mentioned this pull request Jul 31, 2025

Error while installing qhull due to cmake version #40506

Open

GMS103 added t: bug p: blocker / 1 labels Jul 31, 2025

by Artal

496831b

GMS103 mentioned this pull request Jul 31, 2025

typo + guava #40519

Merged

tobiasdiez removed the s: positive review label Aug 1, 2025

conftest.py

fe5ba25

github-actions bot added the s: needs review label Aug 1, 2025

dimpase approved these changes Aug 1, 2025

View reviewed changes

dimpase added s: positive review and removed s: needs review labels Aug 1, 2025

45019 new

2e24b7e

github-actions bot added s: needs review and removed s: positive review labels Aug 2, 2025

45019 new

eccbb0f

vbraun closed this Aug 3, 2025

github-actions bot removed the s: needs review label Aug 3, 2025

Uh oh!

SageMath 10.7 on macOS 15.5 (again) #40512

SageMath 10.7 on macOS 15.5 (again) #40512

Uh oh!

Conversation

GMS103 commented Jul 30, 2025

Uh oh!

GMS103 commented Jul 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dimpase commented Jul 30, 2025

Uh oh!

GMS103 commented Jul 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dimpase commented Jul 30, 2025

Uh oh!

GMS103 commented Jul 30, 2025

Uh oh!

dimpase commented Jul 30, 2025

Uh oh!

GMS103 commented Jul 30, 2025

Uh oh!

dimpase commented Jul 31, 2025

Uh oh!

enriqueartal commented Jul 31, 2025

Uh oh!

GMS103 commented Jul 31, 2025

Uh oh!

GMS103 commented Jul 31, 2025

Uh oh!

GMS103 commented Jul 31, 2025

Uh oh!

enriqueartal commented Jul 31, 2025

Uh oh!

dimpase commented Jul 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dimpase commented Jul 31, 2025

Uh oh!

dimpase commented Jul 31, 2025

Uh oh!

GMS103 commented Jul 31, 2025

Uh oh!

tobiasdiez commented Aug 1, 2025

Uh oh!

GMS103 commented Aug 1, 2025

Uh oh!

dimpase commented Aug 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

GMS103 commented Aug 1, 2025

Uh oh!

GMS103 commented Aug 1, 2025

Uh oh!

vbraun commented Aug 2, 2025

Uh oh!

enriqueartal commented Aug 2, 2025

Uh oh!

vbraun commented Aug 2, 2025

Uh oh!

dimpase commented Aug 2, 2025

Uh oh!

dimpase commented Aug 2, 2025

Uh oh!

GMS103 commented Aug 2, 2025

Uh oh!

github-actions bot commented Aug 2, 2025

Uh oh!

vbraun commented Aug 3, 2025

Uh oh!

GMS103 commented Aug 3, 2025

Uh oh!

Uh oh!

GMS103 commented Jul 30, 2025 •

edited

Loading

GMS103 commented Jul 30, 2025 •

edited

Loading

dimpase commented Jul 31, 2025 •

edited

Loading

dimpase commented Aug 1, 2025 •

edited

Loading