Strange behaviour with doctest

[luisfe]

Hi,

I have recently upgraded my "stable" version of sage to 4.5.3 from
4.51 using
"sage -upgrade"

I upgraded sage with a user called "Alice" having write and read
access to the sources.

I have another user "Bob" that can only read the sources and run sage.
But Bob has no write permissions.

After compiling and building the documentation with Alice, I made a
doctest using Bob:


sage -testall -long

I got the following failures:

sage -t -long "devel/sage/sage/interfaces/qepcad.py"
sage -t -long "devel/sage/sage/schemes/elliptic_curves/
ell_rational_field.py"
sage -t -long "devel/sage/sage/modular/hecke/submodule.py"
sage -t -long "devel/sage/sage/modular/abvar/abvar.py"
sage -t -long "devel/sage/sage/lfunctions/sympow.py"

- qepcad fails with Bob but success with Alice. Some write problems
with temporary files.

- schemes/elliptic_curves/ell_rational_field.py failed consistently
using Bob. With several errors of type:

"RuntimeError: failed to compute analytic rank"

Running the doctest in ell_rational_field with Alice, all doctest
passes and, after running the doctest with Alice once. Bob also
success in running the doctest.

I could not reproduce the failing of the doctest with hecke/
submodule.py and abvar/abvar.py

With /lfunctions/sympow.py happened the same. Bob fails at all
doctest. After running once the doctest with Alice, Bob success.

I am not sure if this is a subtle bug or this behavior is expected/
admissible in some corner cases. It seems like there are parts of sage
that need to be "initialized" at least once by a user with write
permissions.

At least, quepcad fail is a bug to me. If it uses temporary files,
they should be in the user sage's temporary folder. I do not have
installed the experimental package quepcad.

[luisfe]

[Simon King]

Hi Luis!

On Sep 21, 1:57 pm, luisfe <lftab...@yahoo.es> wrote:
> Running the doctest in ell_rational_field with Alice, all doctest
> passes and, after running the doctest with Alice once. Bob also
> success in running the doctest.

So, apparently there is a database that can be read by everybody, but
some computations require write access *to that database*. I would
expect that either as a fail safe (or even as a default), new data are
created in SAGE_TMP, which is in the user's home directory.

> I am not sure if this is a subtle bug or this behavior is expected/
> admissible in some corner cases.

IMHO, it is not admissible.

Cheers,
Simon

[Mitesh Patel]

I've opened

http://trac.sagemath.org/sage_trac/ticket/9965

for the non-qepcad failures.

Could you give the output of the qepcad.py test?

[luisfe]

There is a related bug in trac

http://trac.sagemath.org/sage_trac/ticket/5155

I have posted there quepcad failures

[luisfe]

On Sep 22, 12:28 am, Mitesh Patel <qed...@gmail.com> wrote:
> On 09/21/2010 07:57 AM, luisfe wrote:
> Could you give the output of the qepcad.py test?