Re: [sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

[Dima Pasechnik]

On Fri, 25 Jun 2021 at 07:12, 'jonatha...@googlemail.com' via sage-devel <
sage-devel@googlegroups.com> wrote:

> I did a large computation on a redhat application server 1 1/2 years ago
> and I was very happy that I could just compile gcc to replace the version
> 4.x. There where a couple of problems though, but all of them could be
> resolved. I can't reproduce this however, as those people have thankfully
> provided gcc 10.2.0 in the meantime.
>

I do not see an effort to maintain the ability (sometimes broken?) to
build/install gcc and python
as a core task for Sagemath. It is a burden on bots, on GH Actions we run,
on people who actually maintain it, on users who try going this route
without thinking.

Spin it out into an alternative to Conda if you so inclined.
Such a separation would make it clean for the user that it is not necessary
in 99% of the cases, that it can be updated independently, etc.

I should have added the rest of Sage “toolchain” to the list.


> It's -1 from my side unless:
> - there is a tool that allows installling gcc (or a tested compiler)
> without root privileges,
> - it's well-documented and whatever the instructions, also a non-advanced
> person can follow it,
> - there is still support for the issues with that (it doesn't have to be
> from our community)
>
> In general, I think it is a good idea to join forces. If there is a
> community/tool that deals with these kind of issues already, we could
> invest our resources in other things.
>
> Sidenote: I also think that administrators in universities etc. should
> provide a decent compiler, but I don't think we are going to change this.
> Our university network provides you a bunch of gcc versions and this makes
> life a lot easier for everyone.
> Matthias Koeppe schrieb am Freitag, 25. Juni 2021 um 03:59:20 UTC+2:
>
>> On Thursday, June 24, 2021 at 4:24:10 PM UTC-7 Michael Orlitzky wrote:
>>
>>> On Thu, 2021-06-24 at 15:54 -0700, Volker Braun wrote:
>>>
>>>
>>> > What about replacing gcc with a script that installs the conda
>>> toolchain
>>>
>>>
>>> > automatically. The distro gcc currently does not work for me, so its
>>> not
>>>
>>>
>>> > THAT rare to get into problems. On the other hard, installing conda is
>>> more
>>>
>>>
>>> > likely to work than compiling gcc from scratch.
>>>
>>>
>>>
>>> What does Conda do that a gcc-10 deb/rpm can't do?
>>>
>>
>> It can be installed by unprivileged users.
>>
>>
>>
>
>
>
>
>
>
>
>
> --
>
>
> 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+unsubscr...@googlegroups.com.
>
>
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/fcb07cc1-4a0c-422b-a7df-05e87215693dn%40googlegroups.com
> 
> .
>
>
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq1LW7%2BtRD61uSGYMytRSF3MRJY%2B3v_KVCydO7XC%2B5tapw%40mail.gmail.com.

Re: [sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

['jonatha...@googlemail.com' via sage-devel]

I did a large computation on a redhat application server 1 1/2 years ago 
and I was very happy that I could just compile gcc to replace the version 
4.x. There where a couple of problems though, but all of them could be 
resolved. I can't reproduce this however, as those people have thankfully 
provided gcc 10.2.0 in the meantime.

It's -1 from my side unless:
- there is a tool that allows installling gcc (or a tested compiler) 
without root privileges,
- it's well-documented and whatever the instructions, also a non-advanced 
person can follow it,
- there is still support for the issues with that (it doesn't have to be 
from our community)

In general, I think it is a good idea to join forces. If there is a 
community/tool that deals with these kind of issues already, we could 
invest our resources in other things.

Sidenote: I also think that administrators in universities etc. should 
provide a decent compiler, but I don't think we are going to change this. 
Our university network provides you a bunch of gcc versions and this makes 
life a lot easier for everyone.
Matthias Koeppe schrieb am Freitag, 25. Juni 2021 um 03:59:20 UTC+2:

> On Thursday, June 24, 2021 at 4:24:10 PM UTC-7 Michael Orlitzky wrote:
>
>> On Thu, 2021-06-24 at 15:54 -0700, Volker Braun wrote: 
>> > What about replacing gcc with a script that installs the conda 
>> toolchain 
>> > automatically. The distro gcc currently does not work for me, so its 
>> not 
>> > THAT rare to get into problems. On the other hard, installing conda is 
>> more 
>> > likely to work than compiling gcc from scratch. 
>>
>> What does Conda do that a gcc-10 deb/rpm can't do?
>>
>
> It can be installed by unprivileged users. 
>
>  
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/fcb07cc1-4a0c-422b-a7df-05e87215693dn%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Matthias Koeppe]

On Friday, March 12, 2021 at 2:45:57 AM UTC-8 erik@gmail.com wrote:

> For what it's worth, I was unable to reproduce on my own machine the 
> problem with building R on Cygwin encountered by Matthias which 
> prompted this discussion. [...]
> Since I can't reproduce the problem (yet, though I'm trying some 
> things) and since I build the Windows binary releases I'm less 
> inclined to think it's a blocker issue. 

 
Still missing: The Windows binary release for Sage 9.3.



 

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/9ef7fd01-99aa-419c-8394-c6579a9e5246n%40googlegroups.com.

Re: [sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

[Matthias Koeppe]

On Thursday, June 24, 2021 at 4:24:10 PM UTC-7 Michael Orlitzky wrote:

> On Thu, 2021-06-24 at 15:54 -0700, Volker Braun wrote: 
> > What about replacing gcc with a script that installs the conda toolchain 
> > automatically. The distro gcc currently does not work for me, so its not 
> > THAT rare to get into problems. On the other hard, installing conda is 
> more 
> > likely to work than compiling gcc from scratch. 
>
> What does Conda do that a gcc-10 deb/rpm can't do?
>

It can be installed by unprivileged users. 

 

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/127e0d7a-e3cb-4b8e-8fd0-625f2a4549f2n%40googlegroups.com.

Re: [sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

[Michael Orlitzky]

On Thu, 2021-06-24 at 15:54 -0700, Volker Braun wrote:
> What about replacing gcc with a script that installs the conda toolchain 
> automatically. The distro gcc currently does not work for me, so its not 
> THAT rare to get into problems. On the other hard, installing conda is more 
> likely to work than compiling gcc from scratch.
> 
> 

What does Conda do that a gcc-10 deb/rpm can't do?

I'm probably spoiled since Gentoo allows you to switch the default "cc"
compiler. Is the problem that e.g. gcc-10 and gcc-11 conflict? Or that
with both installed, the user has to specify CC/CXX?


-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/9e2857d269fb811235a0bd0feb83ac59a5e0558c.camel%40orlitzky.com.

[sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

[Volker Braun]

What about replacing gcc with a script that installs the conda toolchain 
automatically. The distro gcc currently does not work for me, so its not 
THAT rare to get into problems. On the other hard, installing conda is more 
likely to work than compiling gcc from scratch.

On Thursday, June 24, 2021 at 12:57:44 PM UTC+2 dim...@gmail.com wrote:

> It's high time we get rid of this annoyances; all the systems Sage
> supports have C/C++/fortran
> compilers capable of building Sage, and Python3 as well.
>
> (and in the very rare cases where it's not available, it's usually a
> problem solved
> by using something like Conda;)
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/763f5e25-ec75-434c-b2dc-978e94457c86n%40googlegroups.com.

[sage-devel] Re: python3-3.9.2 failed to build while running 'sage -i pynormaliz' (ModuleNotFoundError: No module named 'binascii')

[Matthias Koeppe]

There's now a proposed solution in the branch of 
https://trac.sagemath.org/ticket/32056, needs review

On Thursday, June 24, 2021 at 10:59:26 AM UTC-7 Matthias Koeppe wrote:

> I have opened https://trac.sagemath.org/ticket/32056 for the mysterious 
> issue with 'binascii'.
>
>
> On Thursday, June 24, 2021 at 9:46:36 AM UTC-7 marc@gmail.com wrote:
>
>> I am seeing the same error with Python 3.9.5 when I attempt to build Sage 
>> 9.4.beta3 from source on macOS 11.4.  The build fails when it attempts to 
>> import distutils, which depends on the binascii module.  Searching through 
>> the python build directory with find indicates that binascii.o does not 
>> get built.  If I run make in the python build directory using my normal 
>> environment then binascii.o is built, as is the extension module 
>> binascii.cpython-39-darwin.so, and the python build succeeds.  But If I 
>> run make in a sage-buildsh subshell then binascii.o does not get built and 
>> the make command fails in the same way as with the build of sage.  So 
>> evidently something in the sage build environment is preventing python from 
>> compiling its binascii module.
>>
>> I am attaching the log file.
>>
>> - Marc
>>
>> On Friday, June 18, 2021 at 3:09:58 PM UTC-5 max...@gmail.com wrote:
>>
>>> Hi!
>>>
>>> I'm trying to install PyNormaliz within freshly installed Sage 9.3 
>>> within freshly installed Ubuntu 20.04.02 LTS.It breaks giving the following 
>>> error:
>>>
>>> ===
>>> LD_LIBRARY_PATH=/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src
>>>  
>>> CC='gcc -pthread' LDSHARED='gcc -pthread -shared 
>>> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
>>> -L/usr/local/SageMath.93/local/lib 
>>> -Wl,-rpath,/usr/local/SageMath.93/local/lib  -L. 
>>> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
>>> -L/usr/local/SageMath.93/local/lib 
>>> -Wl,-rpath,/usr/local/SageMath.93/local/lib   ' OPT='-DNDEBUG -g -fwrapv 
>>> -O3 -Wall' _TCLTK_INCLUDES='' _TCLTK_LIBS=''   ./python -E 
>>> ./setup.py  build
>>> Traceback (most recent call last):
>>>   File 
>>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/./setup.py",
>>>  
>>> line 33, in 
>>> from distutils import log
>>>   File 
>>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/_distutils_hack/__init__.py",
>>>  
>>> line 83, in create_module
>>> return importlib.import_module('setuptools._distutils')
>>>   File 
>>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/importlib/__init__.py",
>>>  
>>> line 127, in import_module
>>> return _bootstrap._gcd_import(name[level:], package, level)
>>>   File 
>>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/__init__.py",
>>>  
>>> line 16, in 
>>> import setuptools.version
>>>   File 
>>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/version.py",
>>>  
>>> line 1, in 
>>> import pkg_resources
>>>   File 
>>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/pkg_resources/__init__.py",
>>>  
>>> line 23, in 
>>> import zipfile
>>>   File 
>>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/zipfile.py",
>>>  
>>> line 6, in 
>>> import binascii
>>> ModuleNotFoundError: No module named 'binascii'
>>> make[3]: *** [Makefile:638: sharedmods] Error 1
>>>
>>> 
>>> Error building python3-3.9.2
>>>
>>> 
>>> ===
>>>
>>> A full log is attached.
>>>
>>> Please advise.
>>>
>>> Thanks,
>>> Max
>>>
>>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/0e692e6f-6ec0-4508-95fb-de96bb75f6efn%40googlegroups.com.

Re: [sage-devel] Bug with subgroup method?

[David Roe]

H.0 is Sage notation for the 1st generator of H.

sage: preparse("H.0")

'H.gen(0)'

As for getting the value, it depends on what kind of element you have.

sage: H.0._exponents

(2,)
sage: H.0.value()

2
sage: H([1])._exponents

(1,)
sage: H([1]).value()
Traceback (most recent call last):
...
AttributeError...

Also remember that the exponent vectors depend on the context: H.0 is an
element of G (inappropriately in my opinion) and is thus expressed in terms
of a generator of G, while H([1]) is in terms of a generator H.
David



On Thu, Jun 24, 2021 at 3:53 PM Mickaël Hamdad 
wrote:

> And Is there a way to list H directly with numerical values ?
>
> Le jeu. 24 juin 2021 à 21:48, Mickaël Hamdad  a
> écrit :
>
>> Thank you very much
>> Just what H.0 mean ?
>>
>> Le jeu. 24 juin 2021 à 10:21, Samuel Lelievre 
>> a écrit :
>>
>>> Also discussed at Ask Sage:
>>>
>>> - Ask Sage question 57703
>>>   Multiplicative group of Zmod(n)
>>>   https://ask.sagemath.org/question/57703
>>>
>>> --
>>> You received this message because you are subscribed to a topic in the
>>> Google Groups "sage-devel" group.
>>> To unsubscribe from this topic, visit
>>> https://groups.google.com/d/topic/sage-devel/D3zdWqswrFo/unsubscribe.
>>> To unsubscribe from this group and all its topics, send an email to
>>> sage-devel+unsubscr...@googlegroups.com.
>>> To view this discussion on the web visit
>>> https://groups.google.com/d/msgid/sage-devel/0989e81f-6497-4f09-8a51-5311c838773an%40googlegroups.com
>>> 
>>> .
>>>
>> --
> 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+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/CAHJeuofvY5btZv%2BT3gWeb4C8hU7aTiNUHrsD%2BdhTDcbKpnn2Xw%40mail.gmail.com
> 
> .
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAChs6_n76QpcABp7uzmMTXJ%3D-pdybPi7uyZQhu%3Dkad1kmZN2sA%40mail.gmail.com.

Re: [sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

[Matthias Koeppe]

On Thursday, June 24, 2021 at 12:24:10 PM UTC-7 Dima Pasechnik wrote:

> clang on Linux is 
> mature enough)
>

As you know, we are currently not testing this at all. 

 

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/83014893-6887-40e8-8a88-457fede3902bn%40googlegroups.com.

Re: [sage-devel] Bug with subgroup method?

[Mickaël Hamdad]

And Is there a way to list H directly with numerical values ?

Le jeu. 24 juin 2021 à 21:48, Mickaël Hamdad  a
écrit :

> Thank you very much
> Just what H.0 mean ?
>
> Le jeu. 24 juin 2021 à 10:21, Samuel Lelievre 
> a écrit :
>
>> Also discussed at Ask Sage:
>>
>> - Ask Sage question 57703
>>   Multiplicative group of Zmod(n)
>>   https://ask.sagemath.org/question/57703
>>
>> --
>> You received this message because you are subscribed to a topic in the
>> Google Groups "sage-devel" group.
>> To unsubscribe from this topic, visit
>> https://groups.google.com/d/topic/sage-devel/D3zdWqswrFo/unsubscribe.
>> To unsubscribe from this group and all its topics, send an email to
>> sage-devel+unsubscr...@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/sage-devel/0989e81f-6497-4f09-8a51-5311c838773an%40googlegroups.com
>> 
>> .
>>
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAHJeuofvY5btZv%2BT3gWeb4C8hU7aTiNUHrsD%2BdhTDcbKpnn2Xw%40mail.gmail.com.

Re: [sage-devel] Bug with subgroup method?

[Mickaël Hamdad]

Thank you very much
Just what H.0 mean ?

Le jeu. 24 juin 2021 à 10:21, Samuel Lelievre  a
écrit :

> Also discussed at Ask Sage:
>
> - Ask Sage question 57703
>   Multiplicative group of Zmod(n)
>   https://ask.sagemath.org/question/57703
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "sage-devel" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/sage-devel/D3zdWqswrFo/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> sage-devel+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/0989e81f-6497-4f09-8a51-5311c838773an%40googlegroups.com
> 
> .
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAHJeuoeZ2kdbTTAJbJydPpsJ5Nc4uM-85-2Yt-uA2O9Ec%2Bbr4g%40mail.gmail.com.

Re: [sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

[Dima Pasechnik]

On Thu, Jun 24, 2021 at 7:16 PM Matthias Koeppe
 wrote:
>
> Strong -1 on this.
> Given the troubles that we have every time that a major gcc version shows up 
> in distributions (we still do not have GCC 11 support 
> -https://trac.sagemath.org/ticket/31786),

This is usually a waste of time to support people trying to build gcc,
sorry. It often just does not work, as an older gcc is often
impossible to build
with a newer one.
Let them use Conda or clang, this is my answer (clang on Linux is
mature enough). Non-crazy distributions also allow installing
non-bleeding-edge gcc.

> and given the trouble that we had most recently with faulty homebrew 
> packaging of python3, this would dramatically limit our options.
> It would also make test tickets for upcoming releases, such as Python 3.10, 
> in https://trac.sagemath.org/ticket/30767, infeasible.


pyenv provides perfectly functioning Python3, many versions to choose
from. Or the official Python (for macOS).
Or Conda, again, is to the rescue.

>
> On Thursday, June 24, 2021 at 3:57:44 AM UTC-7 Dima Pasechnik wrote:
>>
>> It's high time we get rid of this annoyances; all the systems Sage
>> supports have C/C++/fortran
>> compilers capable of building Sage, and Python3 as well.
>>
>> (and in the very rare cases where it's not available, it's usually a
>> problem solved
>> by using something like Conda;)
>
> --
> 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+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/c2ca9b40-d54c-476d-afdf-17a48b78a535n%40googlegroups.com.

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq1eMSN9F_Pe-eH%3D2NU2eS09LpLOMr85GDNR7BTGKrOUUg%40mail.gmail.com.

Re: [sage-devel] Re: python3-3.9.2 failed to build while running 'sage -i pynormaliz' (ModuleNotFoundError: No module named 'binascii')

[Dima Pasechnik]

On Thu, Jun 24, 2021 at 6:56 PM Marc Culler  wrote:
>
> On Thu, Jun 24, 2021 at 11:50 AM Dima Pasechnik  wrote:
> >
> > There is no need to build Python, you can either use one from Homebrew
> > or from the official Python distribution.
> >
>
> That is an extremely unhelpful answer, especially given what I am
> trying to do.  It is also ridiculous.  If people are following the
> build instructions in the README file and the build is failing then
> the build is broken and it should be fixed.

We have 10 people per day asking what to do with broken Python build, you know,
it is tiring...
Please see
https://doc.sagemath.org/html/en/installation/source.html#linux-recommended-installation

>
> Fortunately, I was able to solve my problem in spite of this unhelpful
> help.  The build is not broken.  But the instructions in the README
> file are incomplete.  The answer which would have helped me is this:
>
> "If you switch branches in the git repository then you need to run
> make distclean before rebuilding sage."
>
> And that really should be added to the README file.
>
> - Marc
>
> PS If you are going to tell people that they do not have to build
> python to build sage then you should also provide instructions for how
> to build sage without building python.

Please see
https://doc.sagemath.org/html/en/installation/source.html#linux-recommended-installation
as well as read the last half a page of the output of ./configure


>
> >
> > > The build fails when it attempts to import distutils, which depends on 
> > > the binascii module.  Searching through the python build directory with 
> > > find indicates that binascii.o does not get built.  If I run make in the 
> > > python build directory using my normal environment then binascii.o is 
> > > built, as is the extension module binascii.cpython-39-darwin.so, and the 
> > > python build succeeds.  But If I run make in a sage-buildsh subshell then 
> > > binascii.o does not get built and the make command fails in the same way 
> > > as with the build of sage.  So evidently something in the sage build 
> > > environment is preventing python from compiling its binascii module.
> > >
> > > I am attaching the log file.
> > >
> > > - Marc
> > >
> > > On Friday, June 18, 2021 at 3:09:58 PM UTC-5 max...@gmail.com wrote:
> > >>
> > >> Hi!
> > >>
> > >> I'm trying to install PyNormaliz within freshly installed Sage 9.3 
> > >> within freshly installed Ubuntu 20.04.02 LTS.It breaks giving the 
> > >> following error:
> > >>
> > >> ===
> > >> LD_LIBRARY_PATH=/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src
> > >>  CC='gcc -pthread' LDSHARED='gcc -pthread -shared 
> > >> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
> > >> -L/usr/local/SageMath.93/local/lib 
> > >> -Wl,-rpath,/usr/local/SageMath.93/local/lib  -L. 
> > >> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
> > >> -L/usr/local/SageMath.93/local/lib 
> > >> -Wl,-rpath,/usr/local/SageMath.93/local/lib   ' OPT='-DNDEBUG -g -fwrapv 
> > >> -O3 -Wall' _TCLTK_INCLUDES='' _TCLTK_LIBS=''   ./python -E 
> > >> ./setup.py  build
> > >> Traceback (most recent call last):
> > >>   File 
> > >> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/./setup.py",
> > >>  line 33, in 
> > >> from distutils import log
> > >>   File 
> > >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/_distutils_hack/__init__.py",
> > >>  line 83, in create_module
> > >> return importlib.import_module('setuptools._distutils')
> > >>   File 
> > >> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/importlib/__init__.py",
> > >>  line 127, in import_module
> > >> return _bootstrap._gcd_import(name[level:], package, level)
> > >>   File 
> > >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/__init__.py",
> > >>  line 16, in 
> > >> import setuptools.version
> > >>   File 
> > >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/version.py",
> > >>  line 1, in 
> > >> import pkg_resources
> > >>   File 
> > >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/pkg_resources/__init__.py",
> > >>  line 23, in 
> > >> import zipfile
> > >>   File 
> > >> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/zipfile.py",
> > >>  line 6, in 
> > >> import binascii
> > >> ModuleNotFoundError: No module named 'binascii'
> > >> make[3]: *** [Makefile:638: sharedmods] Error 1
> > >> 
> > >> Error building python3-3.9.2
> > >> 
> > >> ===
> > >>
> > >> A full log is attached.
> > >>
> > >> Please advise.
> > >>
> > >> Thanks,
> > >> Max
> > >
> > > --
> > > 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 

Re: [sage-devel] proposal - remove gcc, gfortran, python building/spkgs

['Justin C. Walker' via sage-devel]

> On Jun 24, 2021, at 03:57 , Dima Pasechnik  wrote:
> 
> It's high time we get rid of this annoyances; all the systems Sage
> supports have C/C++/fortran
> compilers capable of building Sage, and Python3 as well.
> 
> (and in the very rare cases where it's not available, it's usually a
> problem solved
> by using something like Conda;)

On the face of it, I am not happy with this.  I do not use any package managers 
on any of my macOS systems, and really don’t want to get into the business of 
dealing with the things that go wrong there.

Unless I am missing something, I am a strong “-1” for this move.

Justin

--
Justin C. Walker, Curmudgeon at Large
Director
Institute for the Enhancement of the Director's income
---
--
They said it couldn't be done, but sometimes,
it doesn't work out that way.
  - Casey Stengel
--



-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/281859ED-16C1-4EE4-B85B-ABC29EEA15BB%40mac.com.

[sage-devel] Re: proposal - remove gcc, gfortran, python building/spkgs

[Matthias Koeppe]

Strong -1 on this.
Given the troubles that we have every time that a major gcc version shows 
up in distributions (we still do not have GCC 11 support 
-https://trac.sagemath.org/ticket/31786), and given the trouble that we had 
most recently with faulty homebrew packaging of python3, this would 
dramatically limit our options.
It would also make test tickets for upcoming releases, such as Python 3.10, 
in https://trac.sagemath.org/ticket/30767, infeasible.

On Thursday, June 24, 2021 at 3:57:44 AM UTC-7 Dima Pasechnik wrote:

> It's high time we get rid of this annoyances; all the systems Sage
> supports have C/C++/fortran
> compilers capable of building Sage, and Python3 as well.
>
> (and in the very rare cases where it's not available, it's usually a
> problem solved
> by using something like Conda;)
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/c2ca9b40-d54c-476d-afdf-17a48b78a535n%40googlegroups.com.

Re: [sage-devel] Re: python3-3.9.2 failed to build while running 'sage -i pynormaliz' (ModuleNotFoundError: No module named 'binascii')

[Matthias Koeppe]

On Thursday, June 24, 2021 at 10:56:37 AM UTC-7 marc@gmail.com wrote:

> If people are following the 
> build instructions in the README file and the build is failing then 
> the build is broken and it should be fixed. 
>
>
I agree 
 

> [...] The answer which would have helped me is this: 
>
> "If you switch branches in the git repository then you need to run 
> make distclean before rebuilding sage."
>

I agree that we need to give better instructions for incremental builds. 
There are long-standing issues in our build system, tracked in  
https://trac.sagemath.org/ticket/29708 and 
https://trac.sagemath.org/ticket/31277, that make a "make distclean" 
necesssary.


If "make distclean" helped resolve the issue with "binascii" and the 
python3 spkg was afterwards successfully built, it suggests that the 
python3 installation picks up site packages from the previous installation 
in SAGE_LOCAL. This is a very specific problem that should definitely be 
fixed separately from the above build system issues.



 

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/16118383-be13-4cf7-b5ed-b5b3cce17921n%40googlegroups.com.

Re: [sage-devel] proposal - remove gcc, gfortran, python building/spkgs

[David Roe]

I'm not that well informed of the consequences, but I'm generally
supportive of removing gcc and (especially) gfortran.  We should point
people at other resources to get a functional compiler if there's an
issue.  I also give Dima's opinions on this a lot of weight, since I've
seen him answer question after question from people struggling to build
Sage.  So +1 from me.

Would this mean that Sage not longer installed our own python and we just
used the system python to run Sage (or require people to install their own
python first if the system python were too old)?  I've found it convenient
to have all the python packages Sage installs be self contained in a folder
in userspace, but I'm open to this change if it makes the build process
smoother.  If we do this, we should add some documentation for
Python's venv for Sage developers (like me) who are used to having things
automatically self contained.
David

On Thu, Jun 24, 2021 at 6:57 AM Dima Pasechnik  wrote:

> It's high time we get rid of this annoyances; all the systems Sage
> supports have C/C++/fortran
> compilers capable of building Sage, and Python3 as well.
>
> (and in the very rare cases where it's not available, it's usually a
> problem solved
> by using something like Conda;)
>
> --
> 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+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/CAAWYfq0c4rn62YdQ-U8m7rCCpqzSmeGh%3DO%2BMg99kVv25jDOJig%40mail.gmail.com
> .
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAChs6_%3D0sSRRHvBjxWg0%2B9MLoPnZYYevA-6AytkyqA_KW9%2B38Q%40mail.gmail.com.

[sage-devel] Re: python3-3.9.2 failed to build while running 'sage -i pynormaliz' (ModuleNotFoundError: No module named 'binascii')

[Matthias Koeppe]

I have opened https://trac.sagemath.org/ticket/32056 for the mysterious 
issue with 'binascii'.


On Thursday, June 24, 2021 at 9:46:36 AM UTC-7 marc@gmail.com wrote:

> I am seeing the same error with Python 3.9.5 when I attempt to build Sage 
> 9.4.beta3 from source on macOS 11.4.  The build fails when it attempts to 
> import distutils, which depends on the binascii module.  Searching through 
> the python build directory with find indicates that binascii.o does not 
> get built.  If I run make in the python build directory using my normal 
> environment then binascii.o is built, as is the extension module 
> binascii.cpython-39-darwin.so, and the python build succeeds.  But If I 
> run make in a sage-buildsh subshell then binascii.o does not get built and 
> the make command fails in the same way as with the build of sage.  So 
> evidently something in the sage build environment is preventing python from 
> compiling its binascii module.
>
> I am attaching the log file.
>
> - Marc
>
> On Friday, June 18, 2021 at 3:09:58 PM UTC-5 max...@gmail.com wrote:
>
>> Hi!
>>
>> I'm trying to install PyNormaliz within freshly installed Sage 9.3 within 
>> freshly installed Ubuntu 20.04.02 LTS.It breaks giving the following error:
>>
>> ===
>> LD_LIBRARY_PATH=/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src
>>  
>> CC='gcc -pthread' LDSHARED='gcc -pthread -shared 
>> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
>> -L/usr/local/SageMath.93/local/lib 
>> -Wl,-rpath,/usr/local/SageMath.93/local/lib  -L. 
>> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
>> -L/usr/local/SageMath.93/local/lib 
>> -Wl,-rpath,/usr/local/SageMath.93/local/lib   ' OPT='-DNDEBUG -g -fwrapv 
>> -O3 -Wall' _TCLTK_INCLUDES='' _TCLTK_LIBS=''   ./python -E 
>> ./setup.py  build
>> Traceback (most recent call last):
>>   File 
>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/./setup.py",
>>  
>> line 33, in 
>> from distutils import log
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/_distutils_hack/__init__.py",
>>  
>> line 83, in create_module
>> return importlib.import_module('setuptools._distutils')
>>   File 
>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/importlib/__init__.py",
>>  
>> line 127, in import_module
>> return _bootstrap._gcd_import(name[level:], package, level)
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/__init__.py",
>>  
>> line 16, in 
>> import setuptools.version
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/version.py",
>>  
>> line 1, in 
>> import pkg_resources
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/pkg_resources/__init__.py",
>>  
>> line 23, in 
>> import zipfile
>>   File 
>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/zipfile.py",
>>  
>> line 6, in 
>> import binascii
>> ModuleNotFoundError: No module named 'binascii'
>> make[3]: *** [Makefile:638: sharedmods] Error 1
>>
>> 
>> Error building python3-3.9.2
>>
>> 
>> ===
>>
>> A full log is attached.
>>
>> Please advise.
>>
>> Thanks,
>> Max
>>
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/0fc51e5c-d7cb-422a-a1fd-73660c41fb74n%40googlegroups.com.

Re: [sage-devel] Re: python3-3.9.2 failed to build while running 'sage -i pynormaliz' (ModuleNotFoundError: No module named 'binascii')

[Marc Culler]

On Thu, Jun 24, 2021 at 11:50 AM Dima Pasechnik  wrote:
>
> There is no need to build Python, you can either use one from Homebrew
> or from the official Python distribution.
>

That is an extremely unhelpful answer, especially given what I am
trying to do.  It is also ridiculous.  If people are following the
build instructions in the README file and the build is failing then
the build is broken and it should be fixed.

Fortunately, I was able to solve my problem in spite of this unhelpful
help.  The build is not broken.  But the instructions in the README
file are incomplete.  The answer which would have helped me is this:

"If you switch branches in the git repository then you need to run
make distclean before rebuilding sage."

And that really should be added to the README file.

- Marc

PS If you are going to tell people that they do not have to build
python to build sage then you should also provide instructions for how
to build sage without building python.

>
> > The build fails when it attempts to import distutils, which depends on the 
> > binascii module.  Searching through the python build directory with find 
> > indicates that binascii.o does not get built.  If I run make in the python 
> > build directory using my normal environment then binascii.o is built, as is 
> > the extension module binascii.cpython-39-darwin.so, and the python build 
> > succeeds.  But If I run make in a sage-buildsh subshell then binascii.o 
> > does not get built and the make command fails in the same way as with the 
> > build of sage.  So evidently something in the sage build environment is 
> > preventing python from compiling its binascii module.
> >
> > I am attaching the log file.
> >
> > - Marc
> >
> > On Friday, June 18, 2021 at 3:09:58 PM UTC-5 max...@gmail.com wrote:
> >>
> >> Hi!
> >>
> >> I'm trying to install PyNormaliz within freshly installed Sage 9.3 within 
> >> freshly installed Ubuntu 20.04.02 LTS.It breaks giving the following error:
> >>
> >> ===
> >> LD_LIBRARY_PATH=/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src
> >>  CC='gcc -pthread' LDSHARED='gcc -pthread -shared 
> >> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
> >> -L/usr/local/SageMath.93/local/lib 
> >> -Wl,-rpath,/usr/local/SageMath.93/local/lib  -L. 
> >> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
> >> -L/usr/local/SageMath.93/local/lib 
> >> -Wl,-rpath,/usr/local/SageMath.93/local/lib   ' OPT='-DNDEBUG -g -fwrapv 
> >> -O3 -Wall' _TCLTK_INCLUDES='' _TCLTK_LIBS=''   ./python -E 
> >> ./setup.py  build
> >> Traceback (most recent call last):
> >>   File 
> >> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/./setup.py",
> >>  line 33, in 
> >> from distutils import log
> >>   File 
> >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/_distutils_hack/__init__.py",
> >>  line 83, in create_module
> >> return importlib.import_module('setuptools._distutils')
> >>   File 
> >> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/importlib/__init__.py",
> >>  line 127, in import_module
> >> return _bootstrap._gcd_import(name[level:], package, level)
> >>   File 
> >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/__init__.py",
> >>  line 16, in 
> >> import setuptools.version
> >>   File 
> >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/version.py",
> >>  line 1, in 
> >> import pkg_resources
> >>   File 
> >> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/pkg_resources/__init__.py",
> >>  line 23, in 
> >> import zipfile
> >>   File 
> >> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/zipfile.py",
> >>  line 6, in 
> >> import binascii
> >> ModuleNotFoundError: No module named 'binascii'
> >> make[3]: *** [Makefile:638: sharedmods] Error 1
> >> 
> >> Error building python3-3.9.2
> >> 
> >> ===
> >>
> >> A full log is attached.
> >>
> >> Please advise.
> >>
> >> Thanks,
> >> Max
> >
> > --
> > 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+unsubscr...@googlegroups.com.
> > To view this discussion on the web visit 
> > https://groups.google.com/d/msgid/sage-devel/f53afd48-abd2-440b-a996-f9950e16ad58n%40googlegroups.com.
>
> --
> You received this message because you are subscribed to a topic in the Google 
> Groups "sage-devel" group.
> To unsubscribe from this topic, visit 
> https://groups.google.com/d/topic/sage-devel/euE9MJnfjAo/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to 
> sage-devel+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sa

Re: [sage-devel] Re: python3-3.9.2 failed to build while running 'sage -i pynormaliz' (ModuleNotFoundError: No module named 'binascii')

[Dima Pasechnik]

On Thu, Jun 24, 2021 at 5:46 PM Marc Culler  wrote:
>
> I am seeing the same error with Python 3.9.5 when I attempt to build Sage 
> 9.4.beta3 from source on macOS 11.4.

There is no need to build Python, you can either use one from Homebrew
or from the official Python distribution.


> The build fails when it attempts to import distutils, which depends on the 
> binascii module.  Searching through the python build directory with find 
> indicates that binascii.o does not get built.  If I run make in the python 
> build directory using my normal environment then binascii.o is built, as is 
> the extension module binascii.cpython-39-darwin.so, and the python build 
> succeeds.  But If I run make in a sage-buildsh subshell then binascii.o does 
> not get built and the make command fails in the same way as with the build of 
> sage.  So evidently something in the sage build environment is preventing 
> python from compiling its binascii module.
>
> I am attaching the log file.
>
> - Marc
>
> On Friday, June 18, 2021 at 3:09:58 PM UTC-5 max...@gmail.com wrote:
>>
>> Hi!
>>
>> I'm trying to install PyNormaliz within freshly installed Sage 9.3 within 
>> freshly installed Ubuntu 20.04.02 LTS.It breaks giving the following error:
>>
>> ===
>> LD_LIBRARY_PATH=/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src
>>  CC='gcc -pthread' LDSHARED='gcc -pthread -shared 
>> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
>> -L/usr/local/SageMath.93/local/lib 
>> -Wl,-rpath,/usr/local/SageMath.93/local/lib  -L. 
>> -Wl,-rpath-link,/usr/local/SageMath.93/local/lib 
>> -L/usr/local/SageMath.93/local/lib 
>> -Wl,-rpath,/usr/local/SageMath.93/local/lib   ' OPT='-DNDEBUG -g -fwrapv -O3 
>> -Wall' _TCLTK_INCLUDES='' _TCLTK_LIBS=''   ./python -E 
>> ./setup.py  build
>> Traceback (most recent call last):
>>   File 
>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/./setup.py",
>>  line 33, in 
>> from distutils import log
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/_distutils_hack/__init__.py",
>>  line 83, in create_module
>> return importlib.import_module('setuptools._distutils')
>>   File 
>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/importlib/__init__.py",
>>  line 127, in import_module
>> return _bootstrap._gcd_import(name[level:], package, level)
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/__init__.py",
>>  line 16, in 
>> import setuptools.version
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/setuptools/version.py",
>>  line 1, in 
>> import pkg_resources
>>   File 
>> "/usr/local/SageMath.93/local/lib/python3.9/site-packages/pkg_resources/__init__.py",
>>  line 23, in 
>> import zipfile
>>   File 
>> "/usr/local/SageMath.93/local/var/tmp/sage/build/python3-3.9.2/src/Lib/zipfile.py",
>>  line 6, in 
>> import binascii
>> ModuleNotFoundError: No module named 'binascii'
>> make[3]: *** [Makefile:638: sharedmods] Error 1
>> 
>> Error building python3-3.9.2
>> 
>> ===
>>
>> A full log is attached.
>>
>> Please advise.
>>
>> Thanks,
>> Max
>
> --
> 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+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/f53afd48-abd2-440b-a996-f9950e16ad58n%40googlegroups.com.

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq1Z%2BJWoupqNMrqx19w8kBh3LctrREVjsCdBnyBbwpXKng%40mail.gmail.com.

Re: [sage-devel] Sage 9.4.beta3 failed to build (gcc-10.3.0)

[Kazuyoshi Furutaka]

2021年6月24日木曜日 19:52:38 UTC+9 dim...@gmail.com:

> > What on earth is the GCC included as SPKG in Sage for??? Is it used to 
> build packages at a later time? 
>
> It includes gfortran, which is not available on macOS from a system. 
> Before we started a systematic unvendoring process in 
> https://trac.sagemath.org/ticket/27330, 
> on macOS gfortran was built each time Sage was built. 
>
> It's time to get rid of the monster package, if you asked me. 
>

Ah!  "Try to use as many system packages as possible"!  What a fascinating 
title!
This is the very reason why I started reading Calcote's "Autotools (2ed.)"!
Many of the SPKGs are already included as RPM packages in Fedora, but they 
are not used!  At first I thought they were left undetected by the build 
system (therefore I started the reading), but it is because the GCC in 
Fedora is too NEW and the dependent packages are built along with older 
GCC...

Yours,
Kazuyoshi

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/ba3d92c7-478e-4954-8567-15e1cb237cb7n%40googlegroups.com.

[sage-devel] proposal - remove gcc, gfortran, python building/spkgs

[Dima Pasechnik]

It's high time we get rid of this annoyances; all the systems Sage
supports have C/C++/fortran
compilers capable of building Sage, and Python3 as well.

(and in the very rare cases where it's not available, it's usually a
problem solved
by using something like Conda;)

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq0c4rn62YdQ-U8m7rCCpqzSmeGh%3DO%2BMg99kVv25jDOJig%40mail.gmail.com.

Re: [sage-devel] Sage 9.4.beta3 failed to build (gcc-10.3.0)

[Dima Pasechnik]

On Thu, Jun 24, 2021 at 11:41 AM Kazuyoshi Furutaka
 wrote:
>
>
>
> 2021年6月24日(木) 18:16 Dima Pasechnik :
>>
>> > The Maxima issue is not yet solved, though...
>> You are talking about using gcc to build Sage, right?
>
>
> Yes.
>
>
>>
>> At least for me, ecl  uses clang if it was built with clang.
>
>
> Strange things are...
>
> I don't think annobin is included in GCC package; it's a plugin.
> The commands in the log at the error imply that the build tried to use GCC on 
> the system but the compiler just built for Sage was used (confirmed in the 
> sage-buildsh).
> There're annobins both for GCC and CLANG.
>
> What on earth is the GCC included as SPKG in Sage for???  Is it used to build 
> packages at a later time?

It includes gfortran, which is not available on macOS from a system.
Before we started a systematic unvendoring process in
https://trac.sagemath.org/ticket/27330,
on macOS gfortran was built each time Sage was built.

It's time to get rid of the monster package, if you asked me.


>
> Well... Before upgrading to Fedora-34, everything was smooth.  (But as I 
> remember it's not so long ago when Sage-GCC was updated to version 10...)
>
> Yours,
> Kazuyoshi
> --
> Kazuyoshi Furutaka
>
> --
> 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+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/CANjERvdV4ZOVv6RqibNnFD9KmTsBVwDuCvqcGTP7q53%3Di08gYQ%40mail.gmail.com.

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq2VSG31omHFRs5qDqjj9oFvF2FNqsVkUwTO47CPb8U4jw%40mail.gmail.com.

Re: [sage-devel] Sage 9.4.beta3 failed to build (gcc-10.3.0)

[Kazuyoshi Furutaka]

2021年6月24日(木) 18:16 Dima Pasechnik :

> > The Maxima issue is not yet solved, though...
> You are talking about using gcc to build Sage, right?
>

Yes.



> At least for me, ecl  uses clang if it was built with clang.
>

Strange things are...

   - I don't think annobin is included in GCC package; it's a plugin.
   - The commands in the log at the error imply that the build tried to use
   GCC on the system but the compiler just built for Sage was used (confirmed
   in the sage-buildsh).
   - There're annobins both for GCC and CLANG.

What on earth is the GCC included as SPKG in Sage for???  Is it used to
build packages at a later time?

Well... Before upgrading to Fedora-34, everything was smooth.  (But as I
remember it's not so long ago when Sage-GCC was updated to version 10...)

Yours,
Kazuyoshi
-- 
Kazuyoshi Furutaka

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CANjERvdV4ZOVv6RqibNnFD9KmTsBVwDuCvqcGTP7q53%3Di08gYQ%40mail.gmail.com.

Re: [sage-devel] Sage 9.4.beta3 failed to build (gcc-10.3.0)

[Dima Pasechnik]

On Thu, Jun 24, 2021 at 9:13 AM Kazuyoshi Furutaka
 wrote:
>
> Yes, with this patch in build/pkgs/pynac/patches/, I finally succeeded in 
> building 9.4.beta3 from git source using clang!
>
> Thanks, Dima-san!
It's a pleasure.
>
> The Maxima issue is not yet solved, though...
You are talking about using gcc to build Sage, right?

At least for me, ecl  uses clang if it was built with clang.

>
> 2021年6月23日水曜日 19:06:52 UTC+9 dim...@gmail.com:
>>
>> On Wed, Jun 23, 2021 at 10:49 AM Kazuyoshi Furutaka
>>  wrote:
>> >
>> >
>> >
>> > 2021年6月23日水曜日 17:37:42 UTC+9 dim...@gmail.com:
>> >>
>> >> On Tue, Jun 22, 2021 at 11:57 PM Volker Braun  wrote:
>> >> >
>> >> > I'm running Fedora 34 and 9.4.beta3 builds fine (x86_64). It did 
>> >> > compile the included gcc.
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > On Tuesday, June 22, 2021 at 11:50:00 PM UTC+2 dim...@gmail.com wrote:
>> >> >>
>> >> >>
>> >> >>
>> >> >> On Tue, 22 Jun 2021, 22:16 Kazuyoshi Furutaka, 
>> >> >>  wrote:
>> >> >>>
>> >> >>> On the other hand, build of gcc-10.3 failed. Then, what to do?
>> >> >>
>> >> >>
>> >> >> use another compiler.
>> >> >> e.g. you can use clang
>> >> >>
>> >> >> dnf install clang
>> >> >>
>> >> >> then run Sages configure as follows
>> >> >>
>> >> >> CC=clang CXX=clang++ ./configure
>> >> >>
>> >> >> and then
>> >> >>
>> >> >> make
>> >> >>
>> >> >>
>> >> >> This configuration is not tested much on Linux, but on macOS this is 
>> >> >> the compiler used in our builds.
>> >>
>> >> such a configuration is able to build Sage on Fedora 32, and only one
>> >> doctest fails.
>> >> It uses clang 10.0.1.
>> >>
>> >
>> > I did the same on Fedora 34 (x86_64).
>> > It did not finish because there were too many errors in building 
>> > pynac-0.7.27.p8...
>> you most probably need https://github.com/pynac/pynac/pull/375
>>
>> diff --git a/ginac/numeric.h b/ginac/numeric.h
>> index d620660..50b2b3d 100644
>> --- a/ginac/numeric.h
>> +++ b/ginac/numeric.h
>> @@ -51,6 +51,7 @@
>> #include "ex.h"
>>
>> #include 
>> +#include 
>> #include 
>> #include 
>> #include 
>>
>> which should be in Sage soon.
>>
>>
>> > (log attached)
>> >
>> > Kazuyoshi
>> >
>> > --
>> > 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 view this discussion on the web visit 
>> > https://groups.google.com/d/msgid/sage-devel/694bbf80-46d6-4603-bebb-7f9a50fbb88fn%40googlegroups.com.
>
> --
> 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+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/86d550e0-8a74-4190-927d-1a7f3f4b8c71n%40googlegroups.com.

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq3tJ-PftNn%2BLwEcbnXgBAguKRr2q7pC7Npirw4fKSbKVw%40mail.gmail.com.

Re: [sage-devel] Bug with subgroup method?

[Samuel Lelievre]

Also discussed at Ask Sage:

- Ask Sage question 57703
  Multiplicative group of Zmod(n)
  https://ask.sagemath.org/question/57703

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/0989e81f-6497-4f09-8a51-5311c838773an%40googlegroups.com.

Re: [sage-devel] Sage 9.4.beta3 failed to build (gcc-10.3.0)

[Kazuyoshi Furutaka]

Yes, with this patch 

 
in build/pkgs/pynac/patches/, I finally succeeded in building 9.4.beta3 
from git source using clang!

Thanks, Dima-san!

The Maxima issue is not yet solved, though...

2021年6月23日水曜日 19:06:52 UTC+9 dim...@gmail.com:

> On Wed, Jun 23, 2021 at 10:49 AM Kazuyoshi Furutaka 
>  wrote: 
> > 
> > 
> > 
> > 2021年6月23日水曜日 17:37:42 UTC+9 dim...@gmail.com: 
> >> 
> >> On Tue, Jun 22, 2021 at 11:57 PM Volker Braun  
> wrote: 
> >> > 
> >> > I'm running Fedora 34 and 9.4.beta3 builds fine (x86_64). It did 
> compile the included gcc. 
> >> > 
> >> > 
> >> > 
> >> > 
> >> > On Tuesday, June 22, 2021 at 11:50:00 PM UTC+2 dim...@gmail.com 
> wrote: 
> >> >> 
> >> >> 
> >> >> 
> >> >> On Tue, 22 Jun 2021, 22:16 Kazuyoshi Furutaka, <
> furutaka@gmail.com> wrote: 
> >> >>> 
> >> >>> On the other hand, build of gcc-10.3 failed. Then, what to do? 
> >> >> 
> >> >> 
> >> >> use another compiler. 
> >> >> e.g. you can use clang 
> >> >> 
> >> >> dnf install clang 
> >> >> 
> >> >> then run Sages configure as follows 
> >> >> 
> >> >> CC=clang CXX=clang++ ./configure 
> >> >> 
> >> >> and then 
> >> >> 
> >> >> make 
> >> >> 
> >> >> 
> >> >> This configuration is not tested much on Linux, but on macOS this is 
> the compiler used in our builds. 
> >> 
> >> such a configuration is able to build Sage on Fedora 32, and only one 
> >> doctest fails. 
> >> It uses clang 10.0.1. 
> >> 
> > 
> > I did the same on Fedora 34 (x86_64). 
> > It did not finish because there were too many errors in building 
> pynac-0.7.27.p8... 
> you most probably need https://github.com/pynac/pynac/pull/375 
>
> diff --git a/ginac/numeric.h b/ginac/numeric.h 
> index d620660..50b2b3d 100644 
> --- a/ginac/numeric.h 
> +++ b/ginac/numeric.h 
> @@ -51,6 +51,7 @@ 
> #include "ex.h" 
>
> #include  
> +#include  
> #include  
> #include  
> #include  
>
> which should be in Sage soon. 
>
>
> > (log attached) 
> > 
> > Kazuyoshi 
> > 
> > -- 
> > 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 view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/694bbf80-46d6-4603-bebb-7f9a50fbb88fn%40googlegroups.com.
>  
>
>

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/86d550e0-8a74-4190-927d-1a7f3f4b8c71n%40googlegroups.com.