Metaticket: Add support for python 3.6+

[ohanar]

---

How to build sage with python3:

make configure
./configure --with-python=3
make

Beware that the second step above should be redone after every make distclean.

---

For progress on making all doctests pass, see #26212 and #28298

In order to support python 3.7, the following needs to be fixed:

1. upgrade setuptools to version 2.0.1 #15510 - upgrade to a newer version of setuptools
2. upgrade rpy2 to 2.3.8 #15511 - upgrade to a newer version of rpy2
3. upgrade sympy to version 0.7.4 #15512 - upgrade to a newer version of sympy
4. Update pycrypto to 2.6.1 #14854 - upgrade to a newer version of pycrypto
5. upgrade networkx to version 1.8.1 #15532 - upgrade to a newer version of networkx
6. fix csage to build against python3 #15537 - fix csage to work with python3
7. use Pillow instead of PIL #15539 - switch from using PIL to Pillow
8. trivial pkg fixes needed for python3 #15540 - trivial python3 fixes to a few spkgs
9. sage-location and sage-download-file are broken with python3 #15541 - fix sage-location and sage-download-file for python3
10. remove sqlalchemy #15593 - remove sqlalchemy
11. remove gdmodule #17591 - remove gdmodule
12. StandardError is not Python 3 compatible #15620 - Stop using StandardError
13. Upgrade cython to version 0.20.1 #15755 - upgrade cython to version 0.20.1
14. Upgrade mpmath to version 0.18 #15807 - upgrade mpmath to version 0.18
15. Meta-ticket for Python 3.6+ preparation (sage library) - stage 1 #15980 - meta-ticket for python3 compatibility of the sage library (stage 1)
16. Meta-ticket for Python 3.6+ preparation (sage library) - stage 2 #16052 - meta-ticket for python3 compatibility of the sage library (stage 2)
17. Metaticket: remove c_lib #17854 - remove c_lib
18. Switch from PolyBoRi to BRiAl #18437 - fix polybori and python 3
19. upgrade to pynac-0.3.9.1 #18537 - fix Pynac at least building against python 3
20. Add a python 3 package #17607 - add python 3 package
21. update sagetex to version 3.2 #27024 - make sagetex python3 compatible

Possibly relevant timeline from PEP 373:

Being the last of the 2.x series, 2.7 will have an extended period of maintenance. The current plan is to support it for at least 10 years from the initial 2.7 release. This means there will be bugfix releases until 2020.

Depends on #15510
Depends on #15511
Depends on #15512
Depends on #15531
Depends on #15532
Depends on #15537
Depends on #15539
Depends on #15540
Depends on #15541
Depends on #15807
Depends on #15980
Depends on #18537
Depends on #16052
Depends on #18437
Depends on #27024

Component: python3

Reviewer: Dima Pasechnik

Issue created by migration from https://trac.sagemath.org/ticket/15530

[ohanar]

Changed dependencies from #15510, #15511, #15512 to #15510, #15511, #15512, #15531

[ohanar]

Changed dependencies from #15510, #15511, #15512, #15531 to #15510, #15511, #15512, #15531, #15532

[ohanar]

Changed dependencies from #15510, #15511, #15512, #15531, #15532 to #15510, #15511, #15512, #15531, #15532, #15537

[ohanar]

Changed dependencies from #15510, #15511, #15512, #15531, #15532, #15537 to #15510, #15511, #15512, #15531, #15532, #15537, #15539

[ohanar]

Changed dependencies from #15510, #15511, #15512, #15531, #15532, #15537, #15539 to #15510, #15511, #15512, #15531, #15532, #15537, #15539, #15540

[ohanar]

Changed dependencies from #15510, #15511, #15512, #15531, #15532, #15537, #15539, #15540 to #15510, #15511, #15512, #15531, #15532, #15537, #15539, #15540, #15541

[tscrim]

comment:8

It looks like everything is done. Now we just need to switch Sage to python 3...

[ohanar]

comment:9

Sorry, I should have put a note that I haven't listed all issues (since I haven't even encountered all the issues yet).

[dimpase]

Changed dependencies from #15510, #15511, #15512, #15531, #15532, #15537, #15539, #15540, #15541, #15807, #15980, #18537, #16052, #18437 to #15510, #15511, #15512, #15531, #15532, #15537, #15539, #15540, #15541, #15807, #15980, #18537, #16052, #18437, #27024

[dimpase]

comment:72

Shouldn't this ticket maintain a branch of with all the positively reviewed py3-related tickets?

[embray]

comment:73

That might be useful: I do something like that manually anyways.

[dimpase]

comment:78

What are we going to do with animation support on Python 3?
It seems it is broken - the only (non-console, not sure about console though) possibility seems to be via jupyter+j(s)mol, and this
has not moved far: jupyter/notebook#1872

See also the recent where people complain about broken animation on 8.9.
https://groups.google.com/d/msg/sage-devel/OepzKdldklU/uf2eRVWuDAAJ

[jhpalmieri]

comment:79

Dima, can you provide more details? For the second paragraph, the sage-devel thread gives an example of something which doesn't work in the legacy Sage notebook (their example starts with the command notebook()). So that's not a Python 3 problem. It also doesn't seem to be discussing animation, but maybe I'm missing something. The problem in the first paragraph is about the Jupyter notebook, and I see the same problem with both Python 2 and Python 3.

[dimpase]

comment:80

yes, the sage-devel thread discusses mostly sagenb, but the question is how to do the job in jupyter instead, and this appears to be a blocker, as this is a serious regression.

Disclaimer: I do not know how to do animations in Sage. If someone who knows looked into this, it would be great.

[jhpalmieri]

comment:81

Okay, it still feels like a Jupyter and/or JMol/threejs issue rather than a Python 3 issue.

[dimpase]

comment:82

it is a Python 3 issue, as one does not have sagenb there, while sagenb (at least before Sage 8.9) was able to do animations.

[kcrisman]

comment:83

Regarding the animation issue, see my tests on the sage-devel thread just now. It does somehow seem to be an integration between sagenb and new j(s)mol thing. As for "how to do animations", all they meant was "make a 3d graphic, and then try to make it movable by clicking the little circle in jmol", not using an actual animation command. I'd encourage a few others to try using viewer='jmol' in Jupyter in the 9.0 series to see what happens.

[mkoeppe]

comment:84

Is anything left to be done here or can the ticket be closed?

[dimpase]

comment:85

see #30053 for a blocker for 3.x, for x<7.

[mkoeppe]

comment:86

Let's close this metaticket to signal the completion of the Python 3 transition.

[dimpase]

comment:87

ok, done

[dimpase]

Reviewer: Dima Pasechnik