Some odd doctest failure

[Simon King]

Hi!

I use sage-3.3, upgraded from 3.2.3-sources with a Singular-3-1-0-beta-
spkg.

make check resulted in a handful of failures, some related with
changes in Singular, but one is (probably) unrelated, namely in R:

File "/home/king/SAGE/devel/sage-3.2.3/devel/sage/sage/interfaces/
r.py", line 549:
sage: r.library('foobar')
Expected:
Traceback (most recent call last):
...
ImportError: there is no package called 'foobar'
Got nothing
File "/home/king/SAGE/devel/sage-3.2.3/devel/sage/sage/interfaces/
r.py", line 835:
sage: r.completions('tes')
Expected:
['testPlatformEquivalence', 'testVirtual']
Got:
['testPlatformEquivalence', 'testPlatformEquivalence',
'testVirtual', 'testVirtual']

-----------

I tried to verify the failure, so, here is my complete Sage session on
x86_64 GNU/Linux, AMD Athlon 64 Processor 3700+:
----------------------------------------------------------------------
| Sage Version 3.3, Release Date: 2009-02-21 |
| Type notebook() for the GUI, and license() for information. |
----------------------------------------------------------------------
sage: r.completions('tes')
['testPlatformEquivalence', 'testVirtual']
sage: r.lib
r.library r.library_dynam r.library_dynam_unload
sage: r.library('foobar')
sage: r.completions('tes')
['testPlatformEquivalence', 'testVirtual']
sage: r.library('foobar')
sage: r.completions('tes')

['testPlatformEquivalence',
'testPlatformEquivalence',
'testVirtual',
'testVirtual']
sage: quit
Exiting SAGE (CPU time 0m0.27s, Wall time 4m8.85s).
Exiting spawned R Interpreter process.

Hence: No error is raised when r.library('foobar') is loaded, and
after a while, r.completions('tes') starts to stutter.

Very similar things happen with sage-3.2 on x86_64 GNU/Linux, Dual
Core AMD Opteron Processor 270. Only difference was that the doubling
of r.completions('tes') already occured after the first call of
r.library('foobar').

I can not verify the error on sage.math: r.library('foobar') does
raise an error, and r.completions('tes') works as expected.

Strange, isn't it?

I think this (or a very similar) failure struck me in previous
versions of Sage as well. Is it already tracked somewhere?

Cheers
Simon