Any suggestion on the ways to solve the present problem ? The only one I 
can think of would be to revert to 10.5.beta7 (by detaching head), `make 
build` then `make`.

Advice ?

Le lundi 28 octobre 2024 à 14:33:39 UTC+1, Dima Pasechnik a écrit :

> The upgrade to GAP 4.13.1 simply missed the necessary libsemigroups bump.
> - they should be done in lockstep.
>
> What happens, I think, is that GAP semigroup package tries instead to
> build its own libsemigroup, but it does not work.
>
>
> On Mon, Oct 28, 2024 at 10:11 AM Emmanuel Charpentier
> <[email protected]> wrote:
> >
> > And, BTW :
> >
> > (sage-buildsh) charpent@SAP5342949:sage-10$ ls $PKG_CONFIG_PATH 
> Singular.pc cblas.pc fflas-ffpack.pc gsl.pc libpolys.pc msolve.pc bdw-gc.pc 
> eclib.pc fplll.pc lapack.pc libsemigroups.pc omalloc.pc blas.pc factory.pc 
> givaro.pc libgap.pc linbox.pc singular_resources.pc
> >
> > HTH,
> >
> > Le lundi 28 octobre 2024 à 11:07:17 UTC+1, Emmanuel Charpentier a écrit :
> >>
> >> No such luck (see previous post for config) :
> >>
> >> Le lundi 28 octobre 2024 à 09:25:27 UTC+1, Dima Pasechnik a écrit :
> >>
> >> I think it is probably a Sage's spkgs' dependincies bug. (you need 
> pkg-config for the packages involved).
> >>
> >> In more detail,
> >> GAP package semigroups's sources carry a vendored version of 
> libsemigroups, but it should not be used. Instead its ./configure should 
> find the external (to it) libsemigroups installation, and use it.
> >>
> >> The build script of gap_packages does deal with -I options for 
> particular packages, it assumes that this is taken care of by the 
> aforementioned ./configure - which in this case should figure out that the 
> correct -I by calling
> >>
> >> pkg-config --cflags libsemigroups
> >>
> >> Can you check if it works? (invoke ./sage --buildsh
> >> and in this shell execute the call above)
> >>
> >> charpent@SAP5342949:/usr/local/sage-10$ ./sage --buildsh Starting 
> subshell with Sage environment variables set. Don't forget to exit when you 
> are done. Beware: * Do not do anything with other copies of Sage on your 
> system. * Do not use this for installing Sage packages using "sage -i" or 
> for running "make" at Sage's root directory. These should be done outside 
> the Sage shell. Bypassing shell configuration files... Note: 
> SAGE_ROOT=/usr/local/sage-10 (sage-buildsh) charpent@SAP5342949:sage-10$ 
> pkg-config --cflags libsemigroup Package libsemigroup was not found in the 
> pkg-config search path. Perhaps you should add the directory containing 
> `libsemigroup.pc' to the PKG_CONFIG_PATH environment variable Package 
> 'libsemigroup', required by 'virtual:world', not found
> >>
> >> However (in another window) :
> >>
> >> charpent@SAP5342949:/usr/local/sage-10$ ./sage -optional | grep -v 
> ot_instal /usr/local/sage-10/src/sage/misc/package.py:118: UserWarning: 
> failed to fetch the version of pkg='auditwheel_or_delocate' at 
> https://pypi.org/pypi/auditwheel_or_delocate/json warnings.warn("failed 
> to fetch the version of pkg={!r} at {}".format(pkg, url)) 
> /usr/local/sage-10/src/sage/misc/package.py:118: UserWarning: failed to 
> fetch the version of pkg='ore_algebra' at 
> https://pypi.org/pypi/ore_algebra/json warnings.warn("failed to fetch the 
> version of pkg={!r} at {}".format(pkg, url)) 
> [package]...............................[latest version] ([version]) 
> build...................................? (1.2.1) 
> cypari2.................................? (2.2.0) 
> dot2tex.................................2.11.3.p0 (2.11.3.p0) 
> fricas..................................1.3.11 (1.3.11) 
> gap_jupyter.............................0.9 (0.9) 
> hatch_fancy_pypi_readme.................? (24.1.0) 
> hatch_vcs...............................? (0.4.0) 
> jmol....................................14.29.52 (14.29.52) 
> jupyter_jsmol...........................2022.1.0 (2022.1.0) 
> jupyter_kernel_gap......................? (0.9) 
> jupyter_kernel_singular.................? (0.9.7) 
> libsemigroups...........................2.3.2 (2.3.2) 
> msolve..................................0.6.5 (0.6.5) 
> pysingular..............................0.9.7 (0.9.7) 
> python3_venv............................? (3.12--usr-bin-python3) 
> python_dateutil.........................? (2.9.0.post0) 
> sagemath................................? (0.0.0) 
> sagemath_standard.......................? (10.5b8) 
> singular_jupyter........................0.9.7 (0.9.7)
> >>
> >> Stuck…
> >>
> >>
> >> In fact, pkg-config should be a dependency of libsemigroups (pkgconf 
> should be in build/pkgs/libsemigroups/dependencies), and of gap_packages 
> (pkgconf should be in build/pkgs/gap_packages/dependencies).
> >>
> >> But it is not -- so this is a Sage bug, potentially being the reason 
> behind your problem.
> >>
> >>
> >>
> >>
> >>
> >> On 28 October 2024 01:49:55 GMT, Marc Culler <[email protected]> 
> wrote:
> >> >I know that libsemigroups is a separate spkg. That is not the issue. In
> >> >fact the missing header file exists in the Sage tree, both in the
> >> >local/include directory and in the gap_packages build directory (which 
> is
> >> >still there because the build failed).
> >> >
> >> >% find . -name TextFlow.hpp
> >> >./local/include/libsemigroups/textflowcpp/TextFlow.hpp
> >> 
> >./local/var/tmp/sage/build/gap_packages-4.13.1/src/pkg/semigroups/libsemigroups/extern/textflowcpp/TextFlow.hpp
> >> >
> >> >The problem is that the build script for the new gap_packages spkg is 
> not
> >> >setting the options needed to make the compiler look in either of those
> >> >places. I would assume that a -I option is missing somewhere. Perhaps a
> >> >patch is needed to add that option in the appropriate place.
> >> >
> >> >Here is the compiler command that generates the error, copied from the 
> log
> >> >file:
> >> >
> >> >[spkg-install] g++ -std=gnu++11 -std=gnu++11 -pthread -O2
> >> >-mmacosx-version-min=11.0 -stdlib=libc++ -fno-common -MQ 
> gen/src/bipart.o
> >> >-MMD -MP -MF gen/src/bipart.d -I./bin/include 
> -I./bin/include/libsemigroups
> >> >-std=gnu++14 -O3 -Igapbind14/include/
> >> >-Ilibsemigroups/extern/fmt-8.0.1/include -Ilibsemigroups/include
> >> >-DFMT_HEADER_ONLY -DNDEBUG -O2 -mmacosx-version-min=11.0 
> -stdlib=libc++ -o
> >> >gen/src/bipart.o -I/private/var/tmp/sage-10.5-current/local/include/gap
> >> >-I/private/var/tmp/sage-10.5-current/local/include -DUSE_GASMAN=1 -c
> >> >src/bipart.cpp
> >> >
> >> >It seems to be looking in libsemigroups/include but it should also be
> >> >looking in libsemigroups/include/extern.
> >> >
> >> >- Marc
> >> >
> >> >On Sun, Oct 27, 2024 at 6:26 PM Dima Pasechnik <[email protected]> 
> wrote:
> >> >
> >> >> libsemigroups is a separate spkg, it is (or it should be) a pre-req 
> for
> >> >> gap_packages
> >> >>
> >> >>
> >> >> On 27 October 2024 22:44:53 GMT, Marc Culler <[email protected]>
> >> >> wrote:
> >> >>
> >> >>> I am seeing a build failure for gap_packages-4.13.1 on macOS 15 
> with M1
> >> >>> CPU. A header file seems to be missing:
> >> >>>
> >> >>> [spkg-install] In file included from src/bipart.cpp:38:
> >> >>> [spkg-install] libsemigroups/include/libsemigroups/report.hpp:42:10:
> >> >>> fatal error: 'textflowcpp/TextFlow.hpp' file not found
> >> >>> [spkg-install] #include "textflowcpp/TextFlow.hpp"
> >> >>> [spkg-install] ^~~~~~~~~~~~~~~~~~~~~~~~~~
> >> >>> [spkg-install] 1 error generated.
> >> >>> [spkg-install] make[5]: *** [gen/src/bipart.o] Error 1
> >> >>>
> >> >>> (I don't seem to have the option of attaching files on this list, 
> so the
> >> >>> full log file is not attached.)
> >> >>>
> >> >>> - Marc
> >> >>>
> >> >>
> >> >
> >
> > --
> > 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 [email protected].
> > To view this discussion visit 
> https://groups.google.com/d/msgid/sage-release/2cc864ad-0150-4225-b94c-853a766c660cn%40googlegroups.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 [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/sage-release/ac84bc71-a2ef-40a0-84f3-409f46bde0e3n%40googlegroups.com.

Reply via email to