On Thursday, September 30, 2021 at 5:09:07 AM UTC+9 John H Palmieri wrote:
> One guess would be that ./configure found what it thought was a usable
> installation of openblas, but when gsl tried to build, that openblas was
> not in fact usable.
>
Perhaps. That is strange. It was "not usable"
On Wed, Sep 29, 2021 at 10:52 PM Isuru Fernando wrote:
> > It looks to be required to run tests for networkx, cvxopt and
> symengine. Will tests for those packages just need to be run after
> installing nose manually?
>
> At least for symengine, we can change to pytest. I suppose other packages
> It looks to be required to run tests for networkx, cvxopt and symengine.
Will tests for those packages just need to be run after installing nose
manually?
At least for symengine, we can change to pytest. I suppose other packages
can too.
Isuru
On Wed, Sep 29, 2021 at 4:49 PM David Roe wrote:
I don't have a strong opinion on whether to drop nose or not, but I was
curious how it was used in Sage. It looks to be required to run tests for
networkx, cvxopt and symengine. Will tests for those packages just need to
be run after installing nose manually?
David
On Wed, Sep 29, 2021 at 5:35 P
nose became nose2: https://github.com/nose-devs/nose2 - which is being
developed.
Whether we need so many different testing frameworks is a question.
I'd say, just remove nose.
On Wed, Sep 29, 2021 at 10:35 PM Isuru Fernando wrote:
> Building the nose package requires 'use_2to3' in setuptools,
Building the nose package requires 'use_2to3' in setuptools, but the latest
setuptools version
has dropped support for 'use_2to3'. We can either run 2to3 manually or drop
the nose
package. Since nose hasn't had an update since 2016 and is unmaintained, I
think it's time
to drop it.
Isuru
--
You
One guess would be that ./configure found what it thought was a usable
installation of openblas, but when gsl tried to build, that openblas was
not in fact usable. You can bypass this by either installing a system
version of gsl or using `./configure --with-system-openblas=no` to force
Sage to
I've just opened https://trac.sagemath.org/ticket/32587 to ugrade GSL to
2.7
On Wed, Sep 29, 2021 at 6:04 PM Lorenz Panny wrote:
>
> Trying to build 9.5beta2 using (system) gsl version 2.7 fails
> while compiling src/sage/ext/interpreters/wrapper_cdf.pyx, as
> the .dat member of the gsl_complex
Trying to build 9.5beta2 using (system) gsl version 2.7 fails
while compiling src/sage/ext/interpreters/wrapper_cdf.pyx, as
the .dat member of the gsl_complex structure no longer exists
unless in legacy mode:
https://git.savannah.gnu.org/cgit/gsl.git/tree/complex/gsl_complex.h?id=89168fb10c0
On Wednesday, September 29, 2021 at 2:29:34 AM UTC-7 vdelecroix wrote:
> I had the exact same trouble building the optional package barvinok
I don't think this is the same problem
>
> that has 3 dependencies : ntl, isl and polylib. Running
> 'sage -i barvinok' gave me an error that shows tha
On Wednesday, September 29, 2021 at 2:03:15 AM UTC-7 Kwankyu Lee wrote:
> I started to build the latest Sage 9.5.beta2 from some old Sage
> installation on Mac. Then it failed [...]
>
When incremental builds fail, use "make distclean" to start from scratch.
--
You received this message becau
On Sunday, September 26, 2021 at 10:44:14 AM UTC-7 Matthias Koeppe wrote:
> *1. Replace use of "from sage.PAC.KAGE.all import ..." by more specific
> imports*
>
> This pattern in our Sage library code blocks modularization.
>
In https://trac.sagemath.org/ticket/32586 I would need help from expe
Dear Can Ozan Oğuz,
On Wed, Sep 29, 2021 at 1:44 PM Can Ozan Oğuz wrote:
> Next summer we will be hosting two major conferences ACA (applications
> of computer algebra) and CASC (computer algebra in scientific computing) in
> İstanbul, at Gebze Techical University(GTU). There is a one week gap
Dear all,
Next summer we will be hosting two major conferences ACA (applications of
computer algebra) and CASC (computer algebra in scientific computing) in
İstanbul, at Gebze Techical University(GTU). There is a one week gap
between the two events and we would love host a SageDays then.
Le mercredi 29 septembre 2021 à 13:13:34 UTC+2, john.c...@gmail.com a
écrit :
> Mine is also 1024, so perhaps it is a common default (I have ubuntu
> 20.04.3 LTS, and the only ulimit thing I set to a non-default value is
> 'ulimit -c 0'). Perhaps the build instructions could mention this as
>
On Wed, 29 Sept 2021 at 11:54, Dima Pasechnik wrote:
>
>
>
> On Wed, Sep 29, 2021 at 11:35 AM Ralf Hemmecke wrote:
>>
>> I seem to have a problem with building sage 9.4 via.
>>
>> MAKE='make -j12' make -j12
>>
>> It stops during the build of dochtml. Obviously because of too many open
>> files.
On Wed, Sep 29, 2021 at 11:35 AM Ralf Hemmecke wrote:
> I seem to have a problem with building sage 9.4 via.
>
> MAKE='make -j12' make -j12
>
> It stops during the build of dochtml. Obviously because of too many open
> files. I'm now trying without "-j12" and it seems to be building nicely.
>
>
I seem to have a problem with building sage 9.4 via.
MAKE='make -j12' make -j12
It stops during the build of dochtml. Obviously because of too many open
files. I'm now trying without "-j12" and it seems to be building nicely.
Anyhow... any suggestion what to change to prevent my problem and st
"sage -i " is deprecated. Please use "make ".
On Wed, Sep 29, 2021 at 10:29 AM Vincent Delecroix <
20100.delecr...@gmail.com> wrote:
> I had the exact same trouble building the optional package barvinok
> that has 3 dependencies : ntl, isl and polylib. Running
> 'sage -i barvinok' gave me an err
I had the exact same trouble building the optional package barvinok
that has 3 dependencies : ntl, isl and polylib. Running
'sage -i barvinok' gave me an error that shows that isl was not built
[barvinok-0.41.1] basis_reduction.c:5:10: fatal error: isl/val_gmp.h: No
such file or directory
[barv
On Wed, Sep 29, 2021 at 10:06 AM Kwankyu Lee wrote:
> My question is: why "make" does not build openblas first and then gsl,
> instead of forcing me to do this manually?
"make" does what ./configure tells it to do. Normally speaking one doesn't
need to build openblas nor gsl if using Homebrew o
My question is: why "make" does not build openblas first and then gsl,
instead of forcing me to do this manually?
On Wednesday, September 29, 2021 at 6:03:15 PM UTC+9 Kwankyu Lee wrote:
> Hi,
>
> I started to build the latest Sage 9.5.beta2 from some old Sage
> installation on Mac. Then it f
Hi,
I started to build the latest Sage 9.5.beta2 from some old Sage
installation on Mac. Then it failed while building "gsl" complaining
missing "openblas" library. I could overcome the problem by issuing "sage
-i openblas". I went through this twice on two Sage installations on the
same machi
Hi,
in https://trac.sagemath.org/ticket/32582, I added code to calculate the
colored Jones polynomial.
I'm looking for someone who would like to review the change .
(Usually I know people I can ask for a review, but since this is in an
area, where I haven't contributed anything before, I don'
24 matches
Mail list logo
