polymake 3.1 package failing install on SageMath 8.7

[stefan...@gmail.com]

Hi,

I'm looking at the sagemath package called "polymake".

I tried to install it in SageMath 8.7 and it pulled polymake 3.1

I'm attaching the log. I believe the relevant portion is this:

reference to an undeclared variable $err_prefix at /home/user/SageMath/local/share/polymake/perllib/Polymake/utils.pl line 198. 

Compilation failed in require at /home/user/SageMath/local/share/polymake/perllib/Polymake.pm line 61.

Compilation failed in require at /home/user/SageMath/local/bin/polymake line 162. 

Most of the build goes through rather well, but at the end, this perl build script seems to fail.

I tried to skim through the referenced files utils.pl (line 198) and Polymake.pm (line 61) and apparently

it implements a custom way of declaring variables through a keyword that's called "declare" which is

not part of the Perl language and is custom-built for polymake's build system.

I wasn't able to figure out the problem, but I'm writing here so others can have a look.

Does this package build for someone else?

Thanks,

Stefan

[jplab]

Hi Stefan,

Polymake had some issues with gcc 7 in the past. I do not know if the following could help you:

https://forum.polymake.org/viewtopic.php?f=10&t=579

The latest version (3.4) of polymake is very close to be in Sage:

https://trac.sagemath.org/ticket/24905

Compilation was tested on macOS Mojave, debian stable and ArchLinux. It would be great if you could test this

branch on your computer and let us know if it worked on your OS (either here or on the ticket discussion page).

Best regards,

Jean-Philippe

[Stefan Petrea]

Hi,

I've followed the trac ticket 24905 and I've been able to succesfully build the polymake extension for Sage.

I'm using SageMath 8.7 (that comes with its own gcc 7.2.0).The OS used was Debian buster (10).

Thanks,

Stefan

--
You received this message because you are subscribed to the Google Groups "sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
To post to this group, send email to sage-...@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/92d3c658-0530-466a-9cee-3fb6a391ee0b%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Stefan W]

Apparently I now SageMath doesn't work anymore. If I try to start the Jupyter kernel for Sage

I now get errors about libntl.so.33 (see the log file attached).

Something I found confusing was why ntl and python even, were built throughout the installation

of polymake. Those two (ntl,python) were already available.

[Dima Pasechnik]

ntl and python should not be built by the polymake installation.
Can you start sage in the terminal?

> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To post to this group, send email to sage-...@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/67c6e5e8-5ae7-4d8d-a5d9-7812c883f685%40googlegroups.com.

[Stefan Petrea]

I haven't tried to start it in the terminal. I generally work with Sage through Jupyter.

Unfortunately I don't have that Sage install anymore. I downloaded a fresh copy.

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq1S4EtRwVY%3D2vFESbTdvpNR0y5bZhMghh9Yuyn0C0bk0A%40mail.gmail.com.