Bug with vars()?

[Jean-Pierre Flori]

Dear all,

My Sage 6.6 install quits after printing a depreciation warning message when I issue:
sage: vars()

Same is true for at least globals().
Any clue on what's going on?

Best,
JP

[leif]

Jean-Pierre Flori wrote:
> My Sage 6.6 install quits after printing a depreciation warning message
> when I issue:
> sage: vars()

"Works" for me:

...
'zeta': zeta,
'zeta__exact': <function zeta__exact at 0x7fa3b8c84b90>,
'zeta_symmetric': <function zeta_symmetric at 0x7fa3bac2f9b0>,
'zeta_zeros': <function zeta_zeros at 0x7fa273519aa0>,
'zetaderiv': zetaderiv}
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-1-ef3ed5b882d8> in <module>()
----> 1 vars()

/data/Sage/release/stable/sage-6.6-gcc-4.9.2-default/local/lib/python2.7/site-packages/IPython/core/displayhook.pyc
in __call__(self, result)
230 self.start_displayhook()
231 self.write_output_prompt()
--> 232 format_dict, md_dict = self.compute_format_data(result)
233 self.update_user_ns(result)
234 self.fill_exec_result(result)

TypeError: 'NoneType' object is not iterable
sage:

(Same in 4.7.beta4, FWIW. globals() doesn't throw.)

> Same is true for at least globals().
> Any clue on what's going on?

No. :-)


-leif

[Volker Braun]

There were some changes to how globals() are handled. Did you run "make", can you show us the output, debugger, ...

[John Cremona]

I also have a 6.6 install which quites badly when I give (as the first
command) vars(). The output is hunderds of lines long and ends with

TypeError Traceback (most recent call last)
<ipython-input-1-ef3ed5b882d8> in <module>()
----> 1 vars()

/usr/local/sage/sage-2/local/lib/python2.7/site-packages/IPython/core/displayhook.pyc
in __call__

(self, result)
230 self.start_displayhook()
231 self.write_output_prompt()
--> 232 format_dict, md_dict = self.compute_format_data(result)
233 self.update_user_ns(result)
234 self.fill_exec_result(result)

TypeError: 'NoneType' object is not iterable

> --
> 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 http://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.

[Volker Braun]

vars() works as documented.

Its a good test whether every object that we define can actually print itself. That caughs up plenty of deprecation warnings and fails for 

sage: notebook

<sagenb.notebook.notebook_object.NotebookObject instance at 0x7fbf8537e758>

---------------------------------------------------------------------------

TypeError                                 Traceback (most recent call last)

<ipython-input-1-608e12875613> in <module>()

----> 1 notebook

/mnt/disk/home/release/Sage/local/lib/python2.7/site-packages/IPython/core/displayhook.pyc in __call__(self, result)

    230             self.start_displayhook()

    231             self.write_output_prompt()

--> 232             format_dict, md_dict = self.compute_format_data(result)

    233             self.update_user_ns(result)

    234             self.fill_exec_result(result)

TypeError: 'NoneType' object is not iterable

[John Cremona]

On 6 May 2015 at 12:49, Volker Braun <vbrau...@gmail.com> wrote:
> vars() works as documented.
>
> Its a good test whether every object that we define can actually print

That sounds like a good thing, but I still do not understand why an
error is raised.

John

[Jean-Pierre Flori]

I unpacked the sage 6.6 tarball and ran make.

I'll have some trouble posting the exact output here as the machine is not connected to the network, but basically:
* I'm on an amd64 Ubuntu 14.04.2 LTS.
* I launch: sage
* I type at Sage's prompt: vars()
* I get:
"""
/blablabla/pretty_print.py:147: DepreciationWarning: this is being removed from the global namespace. Use crystals.<tab> instead
See http://trac.sagemath.org/15882 for details.
  ok = representation(obj, self, cycle)
"""
* I'm back at my shell.
I'll have to build gdb to get further info.

With 6.2 and 6.3 and 6.5, it "works" but before everything gets printed I see lots of this warning with slight variations in the line causing it.
With 6.1.1 no warning, I guess 15882 was merged later.

Anyway, it was a false alarm.
I deleted my .sage directory and now it works: I get an error as others do.

[Volker Braun]

Technically its because 

* notebook is a lazy import, 

* printing notebook causes sagenb to be imported

* which changes the displayhook to the SageNB displayhook which causes various problems.

[Jeroen Demeyer]

On 2015-05-06 14:09, Volker Braun wrote:
> Technically its because
> * notebook is a lazy import,
> * printing notebook causes sagenb to be imported
> * which changes the displayhook to the SageNB displayhook which causes
> various problems.

I guess that's a bug. Simply *importing* sagenb should not change the
displayhook.

[Volker Braun]

I opened a ticket for the underlying issue:

==============================================================================

Trac #18512: Move notebook() into Sage

Right now the `notebook()` function is a lazy import from sagenb. Hence any

access (like looking at the docstring) imports sagenb which changes the

display backend which causes various problems (see

https://groups.google.com/d/msg/sage-devel/3uaY6xHLMqQ/Zyn67xJ0LaMJ). It also

doesn't fit into a world where there is more than one notebook. And its

current behavior of turning the current Sage session into a notebook server

can't work with the IPython notebook or SMC.

First, we should definitely move `notebook` into the Sage library to move it

away from sagenb.

The more general question is what to do with it then. Options are

* Just let it launch SageNB, any other notebook needs to be launched via `sage

--notebook=...` on the commandline

* Deprecate `notebook()` altogether

* Let `notebook()` always spawn an independent process, and add an argument to

select which one

Status: new                             Component: notebook

Last modified: 2015-05-26 07:04:55      Created: 2015-05-26 07:04:55 UTC

Report upstream: N/A

Authors:

Reviewers:

Branch:

Keywords:

Dependencies:

------------------------------------------------------------------------------

URL: http://trac.sagemath.org/18512

==============================================================================

On Wednesday, May 6, 2015 at 10:51:27 AM UTC+2, Jean-Pierre Flori wrote: