Why do you have 355 commits on public/experimental_python3_v0?! It is hard to find its way...

On 20/04/2017 08:52, Frédéric Chapoton wrote:
Salut,

you have to apply first : public/experimental_python3_v0

then the ticket 22775 (new conway_polynomial
package) https://trac.sagemath.org/ticket/22775

Frederic

Le jeudi 20 avril 2017 08:40:33 UTC+2, vdelecroix a écrit :

Salut Frédéric,

Then this is the problem. Cython should generate .so files for each .pyx
present in the Sage source code (which should be in the site-packages of
Python3). If you provide an explicit branch with your "hacks" to get
Sage compiled I might be able to provide more help.

Vincent

On 20/04/2017 08:35, Frédéric Chapoton wrote:
Salut,

There is no lazy_import.so in
local/lib/python3.5/site-packages/sage/misc
folder (and in fact no .so file at all there)

And I get

Python 3.5.1 (default, Apr 16 2017, 11:35:45)
Type "copyright", "credits" or "license" for more information.


IPython 5.1.0 -- An enhanced Interactive Python.
?         -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help      -> Python's own help system.
object?   -> Details about 'object', use 'object??' for extra details.


In [1]: import sage.misc.lazy_import

---------------------------------------------------------------------------
ImportError                               Traceback (most recent call
last)
<ipython-input-1-99e0dfbb667e> in <module>()
----> 1 import sage.misc.lazy_import


ImportError: No module named 'sage.misc.lazy_import'


but all .py modules are there:
In [1]: from sage.
               sage.algebras        sage.calculus
 sage.data_structures

               sage.all             sage.categories      sage.databases

               sage.all_cmdline     sage.coding          sage.docs
   >
               sage.all_notebook    sage.combinat        sage.doctest

               sage.arith           sage.crypto          sage.dynamics





Le jeudi 20 avril 2017 08:16:40 UTC+2, vdelecroix a écrit :

Do you have the corresponding (cython) library


SAGE_ROOT/local/lib/python3.5/site-packages/sage/misc/lazy_import.so

If so could you try (inside a python3 shell)

     import sage.misc.lazy_import

that should give you a more precise error.

Vincent

On 19/04/2017 14:57, Frédéric Chapoton wrote:
Here it is  (once again, the failure is related to a cython module):

(sage-sh) chapoton@pc-chapoton:sage3$ python3
Python 3.5.1 (default, Apr  7 2017, 12:17:26)
[GCC 5.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
import sage.all
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File
"/home/chapoton/sage3/local/lib/python3.5/site-packages/sage/all.py",
line 86, in <module>
    import sage.misc.lazy_import
ImportError: No module named 'sage.misc.lazy_import'



Le mercredi 19 avril 2017 13:53:20 UTC+2, vdelecroix a écrit :

Indeed, could you try

   $ sage -sh
   (sage-sh) $ python3
   >>> import sage.all

Vincent

On 19/04/2017 13:11, Frédéric Chapoton wrote:
Hello,

I get nothing from "ls local/lib/python2.7/site-packages/sa*"

and what looks like a correct full sage
in local/lib/python3.5/site-packages/sage

The problem may come from "./sage" calling a python2.7 version ?

Frederic

Le mercredi 19 avril 2017 10:58:53 UTC+2, vdelecroix a écrit :

Hi Frédéric,

Can you check what are inside
local/lib/python3.5/site-packages/sage

   SAGE_ROOT/local/lib/python2.7/site-packages/sage
   SAGE_ROOT/local/lib/python3.5/site-packages/sage

Vincent

On 18/04/2017 21:25, Frédéric Chapoton wrote:
Dear all,

I just managed (using tricks to avoid a few remaining issues in
cythonization and packages) to get make succeed in building sage
with
SAGE_PYTHON3=yes
Sadly, but not unexpectedly, sage then crashes at startup.
Apparently,
it
cannot not find/import any cython extension..

Frédéric

Here is a snapshot of the end of compilation and tentative run:

real 61m21.708s
user 57m46.564s
sys 1m30.508s
Sage build/upgrade complete!
chapoton@icj-laptop:~/sage3$ ./sage

┌────────────────────────────────────────────────────────────────────┐
│ SageMath version 8.0.beta2, Release Date: 2017-04-12
│ Type "notebook()" for the browser-based notebook interface.
│ Type "help()" for help.

└────────────────────────────────────────────────────────────────────┘

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ Warning: this is a prerelease version, and it may be unstable.

┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛
Forcing sage-location, probably because a new package was
installed.
Updating various hardcoded paths...
(Please wait at most a few minutes.)
DO NOT INTERRUPT THIS.
Done updating paths.



**********************************************************************


Oops, Sage crashed. We do our best to make it stable, but...


For the curious people, the crash log ends with

---> 39 from sage.structure.sage_object import SageObject
        global sage.structure.sage_object = undefined
        global SageObject = undefined
     40 from sage.repl.rich_output.output_basic import (
     41     OutputPlainText, OutputAsciiArt, OutputUnicodeArt,
OutputLatex,
     42 )
     43 from sage.repl.rich_output.preferences import
DisplayPreferences
     44
     45
     46 class DisplayException(Exception):
     47     """
     48     Base exception for all rich output-related exceptions.
     49
     50     EXAMPLES::
     51
     52         sage: from sage.repl.rich_output.display_manager
import
DisplayException
     53         sage: raise DisplayException('foo')
     54         Traceback (most recent call last):


ImportError: No module named 'sage.structure.sage_object'










--
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 post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Reply via email to