Sage 3.4.1.rc3 released

[mabshoff]

Hi,

this is rc3 and it is much larger than I anticipated. Loads of
doctests merged, i.e. we are at

Overall weighted coverage score: 68.2%
Total number of functions: 22946
We need 403 more function to get to 70% coverage.
We need 1550 more function to get to 75% coverage.
We need 2698 more function to get to 80% coverage.

Various patches from combinat, John Palmieri's simplicial complexes
and their homology, Solaris doctest fixes by William and me and on and
on.

But rc3 was already so l large and overdue so that I wanted to get
something out for people to test and rebase patches against. The list
of 3.4.1 blockers has gotten much shorter considering that various
blockers were fixed in rc3 that were never listed for rc2:

* #5219: Build ATLAS in dist mode with SSE2 only
* #5559: roots issue on fedora core 32-bit and gcc 4.3.3
* #5662: Gonzalo Tonoria: [with patch, needs review] Timing issue in
clisp-readline (Nasty hang (deadlock?) in maxima pexpect interface on
core 2 quad
* #5697: downgrade GAP due to Itanium issues
* #5741: Detect Atom CPUs as Core2 in the ATLAS detection script
* #5742: ATLAS.spkg: parallel make breaks on system with "real" sh
* #5788: Fix MPIR build problem on OSX 10.4 exposed by linbox

If you look into trac you will note that certain patches have not been
merged despite "positive review" and that is due to me avoiding the
riskier patches. My main goal for tonight, i.e. 18 hours from now is
to resolve the outstanding issues and cut rc4 with everything above
resolved so 3.4.1 can finally be released.

The bits are as usual in

http://sage.math.washington.edu/home/mabshoff/release-cycles-3.4.1/

Please build, doctest and report problems. Note that some build issues
like on OSX 10.4 have not been resolved yet. I have build on 64 bit
linux and 32 bit OSX 10.5 and all test did pass for -long, so I am
hopeful that this release candidate will be reasonably stable.

Cheers,

Michael

Merged in Sage 3.4.1.rc3:

#3081: Kilian Kilger: Added support for Kloosterman sums [Reviewed by
William Stein]
#3309: Dan Drake: massively optimize the binomial function when an
input is real or complex floating point [Reviewed by William Stein]
#4637: David Roe: bug/stupid design of padic_printing.sep print mode
stuff (depends on #5499) [Reviewed by Nick Alexander]
#4724: Nick Alexander: expose pari's galois and finer number field
interfaces [Reviewed by Alex Ghitza]
#5264: Robert Bradshaw: optimize small permgroup elements [Reviewed by
Robert Miller]
#5270: Alex Ghitza: implement plotting of affine and projective
algebraic curves [Reviewed by John Cremona]
#5379: John Palmieri: html conversion problems with live version of
tutorial (ReST version) [Reviewed by William Stein]®
#5386: John Palmieri: simplicial complexes and their homology
[Reviewed by William Stein]
#5450: Jason Grout: plotting a vector as a point plots a sphere, not a
point [Reviewed by Robert Bradshaw]
#5478: Dan Drake: RestrictedPartitions is very slow and a huge memory
hog [Reviewed by Florent Hivert]
#5487: Andrew Mathas: Content function for tableaux [Reviewed by
Florent Hivert, Nicolas Thiery]
#5499: David Roe: Wrong precision when creating p-adic field element
[Reviewed by William Stein]
#5538: Florent Hivert: Family does copy its input + various
improvement [Reviewed by Nicolas Thiery, Michael Abshoff]
#5543: Carl Witty: RealIntervalField parents are not unique [Reviewed
by William Stein]
#5611: John Palmieri: blackboard bold font for jsMath [Reviewed by
Jason Grout]
#5631: Alex Ghitza: improve doctest coverage for schemes/generic/
affine_space.py [Reviewed by William Stein]
#5654: Andrey Novoseltsev: Boundary/interior points of vertices were
computed wrong [Reviewed by Marshall Hampton]
#5655: Andrey Novoseltsev: Improved enumeration of vertices and 0-
dimensional faces [Reviewed by Marshall Hampton]
#5659: Wilfried Huss: Use CRT to speed up solve_mod [Reviewed by
Wilfried Huss]
#5673: John Cremona: Enhanced handling of elliptic curve twists
[Reviewed by Robert Bradshaw]
#5676: Tom Boothby: Preparser identifies hex literals as floating
point. [Reviewed by Robert Bradshaw]
#5684: Willem Jan Palenstijn: Taking negative powers of Laurent
polynomial throws exception [Reviewed by William Stein]
#5685: John Cremona: enhanced nth_root in ZZ and QQ and related
utilities [Reviewed by William Stein]
#5702: Michael Abshoff: make "make check" & friends doctest the
documentation too [Reviewed by Craig Citro]
#5715: Robert Bradshaw: show subdivisions for matrices over GF(2)
[Reviewed by John Palmieri]
#5721: Mike Hansen, Dan Bump: speed up
irreducible_character_freudenthal [Reviewed by Dan Bump, Mike Hansen]
#5733: Robert Miller: bug in 3d plotting of graphs [Reviewed by
William Stein]
#5734: John Cremona: Bring doctests of modular/modsym/manin_symbols.py
up to 100% [Reviewed by David Loeffler]
#5735: Robert Bradshaw: delete extended rationals and integers
completely [Reviewed by William Stein]
#5737: Burcin Erocal: pynac interface improvements [Reviewed by Nick
Alexander]
#5743: Michael Abshoff: Solaris 10/Sparc: Fix numerical noise issues
in doctests [Reviewed by Craig Citro]
#5745: Alex Ghitza: improve doctest coverage for schemes/generic/
projective_space.py [Reviewed by William Stein]
#5747: Alex Ghitza: improve doctest coverage for schemes/generic/
ambient_space.py [Reviewed by William Stein]
#5748: Michael Abshoff: Sage 3.4.1.rc2: isinf() related doctest
failure in sage/rings/infinity.py [Reviewed by Carl Witty]
#5749: Craig Citro: Bring coverage of sage/rings/power_series_poly.pyx
to 100% [Reviewed by William Stein]
#5750: Craig Citro: Bring coverage of sage/matrix/
matrix_integer_2x2.pyx to 100% [Reviewed by Robert Bradshaw]
#5752: Rob Beezer: Bring doctests of sage/games/sudoku.py to 100%
[Reviewed by William Stein]
#5754: John Palmieri: docstrings for all the interactive_constructors
functions are *all* now completely broken [Reviewed by William Stein]
#5756: William Stein: improve coverage of rings/morphism.pyx (and fix
5 bugs in morphism.pyx [Reviewed by Robert Bradshaw]
#5757: John Palmieri: change nodoctest directive [Reviewed by William
Stein]
#5760: Robert Bradshaw: bring doctest coverage of plot3d/shapes.pyx to
100% [Reviewed by William Stein]
#5764: John Palmieri: improve doctest coverage for sageinspect.py
[Reviewed by WIlliam Stein]
#5770: John Cremona: Bring doctests of modular/modsym/p1list.py up to
100% [Reviewed by David Loeffler]
#5772: William Stein: notebook -- typo in twist.py causes Internal
Server Error [Reviewed by Dan Drake]
#5779: Robert Bradshaw: _fast_floats_'s pow returns garbage for non-
integral powers left of zero [Reviewed by Michael Abshoff]
#5780: William Stein: plotting -- deal with NaN's in plot range
[Reviewed by Michael Abshoff]
#5781: Florent Hivert: The empty standard tableau exists [Reviewed by
Nicolas Thiery]
#5782: Robert Bradshaw: Failure of __pow__ in RDF for noninteger
powers of zero [Reviewed by Michael Abshoff]s
#5784: Mitesh Patel: CSS typos [Reviewed by Jason Grout]

[kcrisman]

> #5659: Wilfried Huss: Use CRT to speed up solve_mod [Reviewed by
> Wilfried Huss]

Actually reviewed by William Stein, of course - just in case anyone
thought we now have self-review :)

- kcrisman

[Marshall Hampton]

I am getting errors building linbox on an Intel mac, 10.4.11. I get
several errors of the form:

/usr/libexec/gcc/i686-apple-darwin8/4.0.1/libtool: internal link edit
command failed

This is while attempting to build 3.4.1.rc3 from the tar file (i.e.
not an upgrade).

-Marshall

[Jaap Spies]

mabshoff wrote:
> Hi,
>
> this is rc3 and it is much larger than I anticipated. Loads of
> doctests merged, i.e. we are at
>

On Fedora 9, 32 bit, after an upgrade:

The following tests failed:

sage -t "devel/sage/sage/modular/dirichlet.py"
sage -t "devel/sage/sage/misc/sagedoc.py"

File "/home/jaap/downloads/sage-3.4.1.rc0/devel/sage/sage/modular/dirichlet.py", line 1044:
sage: e.kloosterman_sum_numerical()
Expected:
7.21644966006e-16 + 1.73205080757*I
Got:
6.66133814775e-16 + 1.73205080757*I
**********************************************************************
1 items had failures:
1 of 8 in __main__.example_29
***Test Failed*** 1 failures.
For whitespace errors, see the file /home/jaap/downloads/sage-3.4.1.rc0/tmp/.doctest_dirichlet.py
[6.8 s]
exit code: 1024

sage -t "devel/sage/sage/misc/sagedoc.py"
**********************************************************************
File "/home/jaap/downloads/sage-3.4.1.rc0/devel/sage/sage/misc/sagedoc.py", line 411:
sage: print "ignore this"; search_doc('this creates a polynomial ring') # random # this function has no output: it just prints a string
Exception raised:
Traceback (most recent call last):
File "/home/jaap/downloads/sage-3.4.1.rc0/local/bin/ncadoctest.py", line 1231, in run_one_test
self.run_one_example(test, example, filename, compileflags)
File "/home/jaap/downloads/sage-3.4.1.rc0/local/bin/sagedoctest.py", line 38, in run_one_example
OrigDocTestRunner.run_one_example(self, test, example, filename, compileflags)
File "/home/jaap/downloads/sage-3.4.1.rc0/local/bin/ncadoctest.py", line 1172, in run_one_example
compileflags, 1) in test.globs
File "<doctest __main__.example_8[2]>", line 1, in <module>
print "ignore this"; search_doc('this creates a polynomial ring') # random # this function has no output: it just prints a string###line 411:
sage: print "ignore this"; search_doc('this creates a polynomial ring') # random # this function has no output: it just prints a string
File "/home/jaap/downloads/sage-3.4.1.rc0/local/lib/python2.5/site-packages/sage/misc/sagedoc.py", line 431, in search_doc
pager()(r)
File "/home/jaap/downloads/sage-3.4.1.rc0/local/lib/python2.5/site-packages/IPython/genutils.py", line 1664, in page
term_flags = termios.tcgetattr(sys.stdout)
TypeError: argument must be an int, or have a fileno() method.
**********************************************************************
1 items had failures:
1 of 3 in __main__.example_8
***Test Failed*** 1 failures.
For whitespace errors, see the file /home/jaap/downloads/sage-3.4.1.rc0/tmp/.doctest_sagedoc.py
[53.3 s]
exit code: 1024

Cheers,

Jaap

[mabshoff]

On Apr 16, 12:21 pm, Marshall Hampton <hampto...@gmail.com> wrote:
> I am getting errors building linbox on an Intel mac, 10.4.11.  I get
> several errors of the form:
>
> /usr/libexec/gcc/i686-apple-darwin8/4.0.1/libtool: internal link edit
> command failed

These are known problems, i.e. from above:

*Note that some build issues like on OSX 10.4 have not been resolved
yet*

and specificially "#5788: Fix MPIR build problem on OSX 10.4 exposed
by linbox "

> This is while attempting to build 3.4.1.rc3 from the tar file (i.e.
> not an upgrade).
>
> -Marshall
>

Cheers,

Michael

[mabshoff]

On Apr 16, 1:02 pm, Jaap Spies <j.sp...@hccnet.nl> wrote:
> mabshoff wrote:
> > Hi,

Hi Jaap,

> > this is rc3 and it is much larger than I anticipated. Loads of
> > doctests merged, i.e. we are at
>
> On Fedora 9, 32 bit, after an upgrade:
>
> The following tests failed:
>
>         sage -t  "devel/sage/sage/modular/dirichlet.py"
>         sage -t  "devel/sage/sage/misc/sagedoc.py"
>
> File "/home/jaap/downloads/sage-3.4.1.rc0/devel/sage/sage/modular/dirichlet.py", line 1044:
>      sage: e.kloosterman_sum_numerical()
> Expected:
>      7.21644966006e-16 + 1.73205080757*I
> Got:
>      6.66133814775e-16 + 1.73205080757*I
> **********************************************************************
> 1 items had failures:
>     1 of   8 in __main__.example_29
> ***Test Failed*** 1 failures.
> For whitespace errors, see the file /home/jaap/downloads/sage-3.4.1.rc0/tmp/.doctest_dirichlet.py
>          [6.8 s]
> exit code: 1024

Please open a ticket for this one. It is numerical noise and trivial
to fix.

Not sure what is happening here. In case it is reproducible for you
please open another ticket.

> Cheers,
>
> Jaap

Cheers,

Michael

[davidloeffler]

On Apr 16, 9:02 pm, Jaap Spies <j.sp...@hccnet.nl> wrote:
> On Fedora 9, 32 bit, after an upgrade:
>
> The following tests failed:
>
> sage -t "devel/sage/sage/modular/dirichlet.py"
> sage -t "devel/sage/sage/misc/sagedoc.py"

On 32bit SuSE Linux, upgrading from 3.4.1.rc2, I also get the
dirichlet.py error (and I get the same different answer that Jaap
gets, 6.66133814775e-16 + 1.73205080757*I). I am not seeing the
sagedoc.py error though.

David

[mabshoff]

Oops, thanks for pointing that out. Fixed.

> - kcrisman

Cheers,

Michael

[Jaap Spies]

mabshoff wrote:
>
>
> On Apr 16, 1:02 pm, Jaap Spies <j.sp...@hccnet.nl> wrote:
>> mabshoff wrote:
>>> Hi,
>

Hi Michael,

>
> Please open a ticket for this one. It is numerical noise and trivial
> to fix.
>

Done.

>> sage -t "devel/sage/sage/misc/sagedoc.py"
>> **********************************************************************
[...]

>> exit code: 1024
>
> Not sure what is happening here. In case it is reproducible for you
> please open another ticket.
>

Not sure what you mean, but this happens all the time repeating
this test.

Jaap

[mabshoff]

On Apr 16, 2:18 pm, Jaap Spies <j.sp...@hccnet.nl> wrote:
> mabshoff wrote:

Hi Jaap,

<SNIP>

> > Not sure what is happening here. In case it is reproducible for you
> > please open another ticket.
>
> Not sure what you mean, but this happens all the time repeating
> this test.

That meant please open a ticket if it is reproducible :)

> Jaap

Cheers,

Michael

[Justin C. Walker]

On Apr 16, 2009, at 09:10 , mabshoff wrote:

>
> Hi,
>
> this is rc3 and it is much larger than I anticipated. Loads of
> doctests merged, i.e. we are at

[snip]

>
> The bits are as usual in
>
> http://sage.math.washington.edu/home/mabshoff/release-cycles-3.4.1/

I built rc3 as an upgrade from rc2, upgrading from a full build of rc1
(Mac OS X, 10.5.6, Dual Quad Xeon). The build succeeded, but I
noticed 10 instances of this:

preparing documents... WARNING: html_favicon is not an .ico file

In addition, when I view, e.g., the tutorial, in Safari, I see these
missing pieces:

file:///Users/tmp/sage-3.4.1.rc1/devel/sage/doc/output/html/en/tutorial/_static/sage_logo.png
file:///javascript_local/jsMath/easy/load.js

Testing ("-j6") tells me:
All tests passed!
Total time for all tests: 3921.3 seconds
However, an ERROR occurred in the Distributed Sage unit tests.

The distributed sage test output shows:

=
=
=
=
=
=
=
========================================================================
[ERROR]:
sage
.dsage
.twisted.tests.test_remote.ClientRemoteCallsTest.testremoteSubmitJob

Traceback (most recent call last):

Failure: twisted.trial.util.DirtyReactorAggregateError: Reactor
was unclean.
Selectables:
<Broker #0 on 51757>

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

I placed test.log and the output from the upgrade in
sage.math.washington.edu:~justin/logs/3.4.1.rc3.{test,upgrade}.log
respectively.

Justin

--
Justin C. Walker
Curmudgeon at Large
Director
Institute for the Enhancement of the Director's Income
--
Build a man a fire and he'll be warm
for a night.
Set a man on fire and he'll be warm
for the rest of his life.

[John Cremona]

On 32-bit ubuntu, after a successful build from scratch:

The following tests failed:
sage -t "devel/sage/sage/misc/sagedoc.py"
sage -t "devel/sage/sage/modular/dirichlet.py"

The first looks like a timeout. The second has been reported already.

File "/home/john/sage-3.4.1.rc3/devel/sage/sage/modular/dirichlet.py",

line 1044:
sage: e.kloosterman_sum_numerical()
Expected:
7.21644966006e-16 + 1.73205080757*I
Got:

5.55111512313e-16 + 1.73205080757*I

John

2009/4/17 Justin C. Walker <jus...@mac.com>:

[mabshoff]

On Apr 17, 12:37 am, John Cremona <john.crem...@gmail.com> wrote:

Hi John,

> On 32-bit ubuntu, after a successful build from scratch:
>
> The following tests failed:
>         sage -t  "devel/sage/sage/misc/sagedoc.py"

This one has actually been reported - see #5806 - but I would be
surprised if it were a timeout. Do you still have the log?

>         sage -t  "devel/sage/sage/modular/dirichlet.py"
>
> The first looks like a timeout.  The second has been reported already.

Yep.

> File "/home/john/sage-3.4.1.rc3/devel/sage/sage/modular/dirichlet.py",
> line 1044:
>     sage: e.kloosterman_sum_numerical()
> Expected:
>     7.21644966006e-16 + 1.73205080757*I
> Got:
>     5.55111512313e-16 + 1.73205080757*I

Hmm, is there any better input to pick for e so that the real value
isn't more or less zero? I can fix the test as numerical noise by
changing it to

'...e-16 + 1.73205080757*I'

but I would prefer to avoid that if a better value can be found.

> John

Cheers,

Michael

[mabshoff]

On Apr 17, 12:27 am, "Justin C. Walker" <jus...@mac.com> wrote:
> On Apr 16, 2009, at 09:10 , mabshoff wrote:
> > Hi,

Hi Justin,

> > this is rc3 and it is much larger than I anticipated. Loads of
> > doctests merged, i.e. we are at
> [snip]
>
> > The bits are as usual in
>
> >  http://sage.math.washington.edu/home/mabshoff/release-cycles-3.4.1/
>
> I built rc3 as an upgrade from rc2, upgrading from a full build of rc1  
> (Mac OS X, 10.5.6, Dual Quad Xeon).  The build succeeded, but I  
> noticed 10 instances of this:
>
>      preparing documents... WARNING: html_favicon is not an .ico file
>
> In addition, when I view, e.g., the tutorial, in Safari, I see these  
> missing pieces:
>
>      file:///Users/tmp/sage-3.4.1.rc1/devel/sage/doc/output/html/en/tutorial/_static/sage_logo.png
>      file:///javascript_local/jsMath/easy/load.js

I think Pat posted a patch to fix this, i.e. #5799. It is quite a nice
patch, so if someone could review it that would be very nice :)

> Testing ("-j6") tells me:
>      All tests passed!
>      Total time for all tests: 3921.3 seconds
>      However, an ERROR occurred in the Distributed Sage unit tests.
>
> The distributed sage test output shows:

<SNIP>

> I placed test.log and the output from the upgrade in
>     sage.math.washington.edu:~justin/logs/3.4.1.rc3.{test,upgrade}.log
> respectively.

Hmm, is this failure reproducible? Funny firewall settings can cause
this and let's just say that DSage isn't exactly the most robust
component in Sage :(

> Justin

Cheers,

Michael

[John Cremona]

2009/4/17 mabshoff <mabs...@googlemail.com>:

>
> On Apr 17, 12:37 am, John Cremona <john.crem...@gmail.com> wrote:
>
> Hi John,
>
>> On 32-bit ubuntu, after a successful build from scratch:
>>
>> The following tests failed:
>> sage -t "devel/sage/sage/misc/sagedoc.py"
>
>
> This one has actually been reported - see #5806 - but I would be
> surprised if it were a timeout. Do you still have the log?

That was on my laptop which I left at home.... so I cannot help for
about 7 hours.

>
>> sage -t "devel/sage/sage/modular/dirichlet.py"
>>
>> The first looks like a timeout. The second has been reported already.
>
> Yep.
>
>> File "/home/john/sage-3.4.1.rc3/devel/sage/sage/modular/dirichlet.py",
>> line 1044:
>> sage: e.kloosterman_sum_numerical()
>> Expected:
>> 7.21644966006e-16 + 1.73205080757*I
>> Got:
>> 5.55111512313e-16 + 1.73205080757*I
>
> Hmm, is there any better input to pick for e so that the real value
> isn't more or less zero? I can fix the test as numerical noise by
> changing it to
>
> '...e-16 + 1.73205080757*I'
>
> but I would prefer to avoid that if a better value can be found.

That would be a good idea, but I don't know enough about Kloosterman
sums to help, sorry.

John

>
>> John
>
> Cheers,
>
> Michael
> >
>

[John Cremona]

On 64-bit Ubuntu (Bill Hart's machine) all tests pass.

John

2009/4/17 John Cremona <john.c...@gmail.com>:

[mabshoff]

On Apr 17, 2:57 am, John Cremona <john.crem...@gmail.com> wrote:
> 2009/4/17 mabshoff <mabsh...@googlemail.com>:

<SNIP>

> >> File "/home/john/sage-3.4.1.rc3/devel/sage/sage/modular/dirichlet.py",
> >> line 1044:
> >>     sage: e.kloosterman_sum_numerical()
> >> Expected:
> >>     7.21644966006e-16 + 1.73205080757*I
> >> Got:
> >>     5.55111512313e-16 + 1.73205080757*I
>
> > Hmm, is there any better input to pick for e so that the real value
> > isn't more or less zero? I can fix the test as numerical noise by
> > changing it to
>
> >  '...e-16 + 1.73205080757*I'
>
> > but I would prefer to avoid that if a better value can be found.
>
> That would be a good idea, but I don't know enough about Kloosterman
> sums to help, sorry.

No problem, on Solaris/x86-64 32 bit I am seeing a slightly different
numerical noise value (this is rc2 upgraded to rc3):

sage -t -long devel/sage/sage/modular/dirichlet.py
**********************************************************************
File "/home/mabshoff/build-3.4.1.rc2/sage-3.4.1.rc2-fulvia-gcc-4.3.3/
devel/sage-main/sage/modular/dirichlet.py", line 10

44:
sage: e.kloosterman_sum_numerical()
Expected:
7.21644966006e-16 + 1.73205080757*I
Got:

6.66133814775e-16 + 1.73205080757*I
**********************************************************************

So I will go with the proposed solution for now.

> John

Cheers,

Michael

[mabshoff]

Ok, after analyzing all reports here, the build farm on boxen and
problems on selcted boxen on SkyNet it seems that we have a total of
four failures introduced in 3.4.1.rc3:

* #5805: Sage 3.4.1.rc3: numerical noise in "devel/sage/sage/modular/
dirichlet.py"
* #5806: Sage 3.4.1.rc3: failing test "devel/sage/sage/misc/
sagedoc.py"
* #5810: Sage 3.4.1.rc3: Fedora 10/64 - unable to start Maxima issue
in shapes.pyx
* #5811: Sage 3.4.1.rc3: Fedora 10/64 - type_reducible.py doctest
failure due to '__cmp__'

#5805 is trivial to fix, #5806 has some leads by Dan Drake. I got a
suspicion on how to fix #5811, but I am clueless about #5810 at the
moment.

Please report any additional failures you see since 3.4.1.rc4 ought to
drop tonight (Seattle time).

Cheers,

Michael

[mabshoff]

On Apr 17, 4:35 am, mabshoff <mabsh...@googlemail.com> wrote:
> Ok, after analyzing all reports here, the build farm on boxen and
> problems on selcted boxen on SkyNet it seems that we have a total of
> four failures introduced in 3.4.1.rc3:
>
>  * #5805: Sage 3.4.1.rc3: numerical noise in "devel/sage/sage/modular/
> dirichlet.py"
>  * #5806: Sage 3.4.1.rc3: failing test "devel/sage/sage/misc/
> sagedoc.py"
>  * #5810: Sage 3.4.1.rc3: Fedora 10/64 - unable to start Maxima issue
> in shapes.pyx
>  * #5811: Sage 3.4.1.rc3: Fedora 10/64 - type_reducible.py doctest
> failure due to '__cmp__'
>
> #5805 is trivial to fix, #5806 has some leads by Dan Drake. I got a
> suspicion on how to fix #5811, but I am clueless about #5810 at the
> moment.

Oops, I am clueless about #5811, but I have an idea about #5810.

There is also a reported "make check" failure for MPIR 1.0.rc8 on a
Pentium D by Bucrin, but MPIR 1.1 does pass make check on that box and
the plan was to update anyway.

> Please report any additional failures you see since 3.4.1.rc4 ought to
> drop tonight (Seattle time).
>
> Cheers,
>
> Michael

Cheers,

Michael

[Pat LeSmithe]

I'm not sure if it helps, but a ticket search yields

http://trac.sagemath.org/sage_trac/ticket/4158

[mabshoff]

On Apr 17, 5:01 am, Pat LeSmithe <qed...@gmail.com> wrote:

Hi,

> I'm not sure if it helps, but a ticket search yields
>
> http://trac.sagemath.org/sage_trac/ticket/4158

Yes, I am aware of that ticket and unfortunately it does not really
help. Last time this specific problem popped up (around Sage 3.1.4) it
also just as mysteriously went away. For various reasons we now have
another chance to fix this once and for all (I had assumed it only
popped up on FC 9/64 bit with gcc 4.3.3 which is not the standard
system compiler), so I am glad it is triggering on another box. The
common factor here is Fedora systems, i.e. I have not been able to
trigger this on RHEL/Centos which ought to be closely related or other
Linux, OSX or Solaris systems, so this ought to be interesting.

Cheers,

Michael

[John H Palmieri]

On Apr 17, 4:35 am, mabshoff <mabsh...@googlemail.com> wrote:

> Ok, after analyzing all reports here, the build farm on boxen and
> problems on selcted boxen on SkyNet it seems that we have a total of
> four failures introduced in 3.4.1.rc3:
>
>  * #5805: Sage 3.4.1.rc3: numerical noise in "devel/sage/sage/modular/
> dirichlet.py"
>  * #5806: Sage 3.4.1.rc3: failing test "devel/sage/sage/misc/
> sagedoc.py"
>  * #5810: Sage 3.4.1.rc3: Fedora 10/64 - unable to start Maxima issue
> in shapes.pyx
>  * #5811: Sage 3.4.1.rc3: Fedora 10/64 - type_reducible.py doctest
> failure due to '__cmp__'
>
> #5805 is trivial to fix, #5806 has some leads by Dan Drake. I got a
> suspicion on how to fix #5811, but I am clueless about #5810 at the
> moment.

For the record, I don't know what's going on with #5806.

On a linux box -- it's running ubuntu, and 'uname -a' tells me

Linux jpalmieri538 2.6.24-23-generic #1 SMP Wed Apr 1 21:47:28 UTC
2009 i686 GNU/Linux

-- I'm seeing two more unreported failures, randstate and sage0:

sage -t -long "devel/sage/sage/misc/randstate.pyx"
**********************************************************************
File "/home/palmieri/Documents/sage-3.4.1.rc3/devel/sage/sage/misc/
randstate.pyx", line 124:
: s = ZZ(subsage('initial_seed()'))
Exception raised:

Traceback (most recent call last):

File "/home/palmieri/Documents/sage-3.4.1.rc3/local/bin/
ncadoctest.py", line 1231, in run_on\

e_test
self.run_one_example(test, example, filename, compileflags)

File "/home/palmieri/Documents/sage-3.4.1.rc3/local/bin/
sagedoctest.py", line 38, in run_one\

_example
OrigDocTestRunner.run_one_example(self, test, example,
filename, compileflags)

File "/home/palmieri/Documents/sage-3.4.1.rc3/local/bin/
ncadoctest.py", line 1172, in run_on\

e_example
compileflags, 1) in test.globs

File "<doctest __main__.example_0[32]>", line 1, in <module>
s = ZZ(subsage('initial_seed()'))###line 124:
: s = ZZ(subsage('initial_seed()'))
File "parent.pyx", line 288, in
sage.structure.parent.Parent.__call__ (sage/structure/parent\
.c:4320)
File "coerce_maps.pyx", line 155, in
sage.structure.coerce_maps.NamedConvertMap._call_ (sage\
/structure/coerce_maps.c:4225)
File "/home/palmieri/Documents/sage-3.4.1.rc3/local/lib/
python2.5/site-packages/sage/interfa\
ces/expect.py", line 1648, in _integer_
return sage.rings.all.Integer(repr(self))
File "integer.pyx", line 517, in
sage.rings.integer.Integer.__init__ (sage/rings/integer.c:6\
446)
TypeError: unable to convert x
(=-------------------------------------------------------------\
--------------
AttributeError Traceback (most recent
call last)

/home/palmieri/.sage/temp/jpalmieri538/32330/
_home_palmieri__sage_init_sage_0.py in <module>()

/home/palmieri/Documents/sage-3.4.1.rc3/local/lib/python2.5/site-
packages/sage/misc/functional\
.pyc in gen(x)
353 Return the generator of x.
354 """
--> 355 return x.gen()
356
357 def gens(x):

AttributeError: 'int' object has no attribute 'gen') to an integer
**********************************************************************
File "/home/palmieri/Documents/sage-3.4.1.rc3/devel/sage/sage/misc/
randstate.pyx", line 131:
: r == ZZ.random_element(2^200)
Expected:
True
Got:
False


sage -t -long "devel/sage/sage/interfaces/sage0.py"
**********************************************************************
File "/home/palmieri/Documents/sage-3.4.1.rc3/devel/sage/sage/
interfaces/sage0.py", line 55:
sage: a^3
Expected:
8
Got:
<BLANKLINE>
**********************************************************************
File "/home/palmieri/Documents/sage-3.4.1.rc3/devel/sage/sage/
interfaces/sage0.py", line 62:
sage: V.gens()
Expected:
((1, 0, 0, 0), (0, 1, 0, 0), (0, 0, 1, 0), (0, 0, 0, 1))
Got:
<BLANKLINE>
**********************************************************************
File "/home/palmieri/Documents/sage-3.4.1.rc3/devel/sage/sage/
interfaces/sage0.py", line 75:
sage: g = V.0; g
Expected:
(1, 0, 0, 0)
Got:
<BLANKLINE>


This one goes on for a while; the full log is at

sage.math.washington.edu/~palmieri/misc/sage0.log

[Justin C. Walker]

I applied the 'ext' patch and the 'alt sage' patch.

On Apr 17, 2009, at 02:52 , mabshoff wrote:
> On Apr 17, 12:27 am, "Justin C. Walker" <jus...@mac.com> wrote:
>> On Apr 16, 2009, at 09:10 , mabshoff wrote:
>>> The bits are as usual in
>>
>>> http://sage.math.washington.edu/home/mabshoff/release-cycles-3.4.1/

[snip]

>>> noticed 10 instances of this:
>>
>> preparing documents... WARNING: html_favicon is not an .ico file

I rebuilt the "en/tuturial html" doc as a test. I still get the above
warning.

>> In addition, when I view, e.g., the tutorial, in Safari, I see these
>> missing pieces:
>>
>> file:///Users/tmp/sage-3.4.1.rc1/devel/sage/doc/output/html/en/tutorial/_static/sage_logo.png
>> file:///javascript_local/jsMath/easy/load.js

These errors no longer occur, so the page loads w/o errors. However,
when I stepped forward in the tutorial, to
<file:///Users/tmp/sage-3.4.1.rc1/devel/sage/doc/output/html/en/tutorial/tour.html
>
I got
file:///Users/tmp/sage-3.4.1.rc1/devel/sage/doc/output/html/en/tutorial/_static/sageicon.png
does not exist

The problem disappeared when I tried to reproduce it, and I got no
repeats. I assume it was due to cacheing, and I'm just reporting this
for completeness.

I gave the ticket two thumbs up :-}

Justin

--
Justin C. Walker, Curmudgeon-At-Large

Institute for the Enhancement of the Director's Income

--------
Experience is what you get
when you don't get what you want.
--------

[Justin C. Walker]

A quick follow-up:

On Apr 17, 2009, at 02:52 , mabshoff wrote:

> On Apr 17, 12:27 am, "Justin C. Walker" <jus...@mac.com> wrote:
>> On Apr 16, 2009, at 09:10 , mabshoff wrote:

[snip]

>> preparing documents... WARNING: html_favicon is not an .ico file

As I mentioned in my last note, this warning still occurs, but since
it is "just a warning", I didn't hold up the patch review (since the
patch didn't claim to address it).

[snip]

>> Testing ("-j6") tells me:
>> All tests passed!
>> Total time for all tests: 3921.3 seconds
>> However, an ERROR occurred in the Distributed Sage unit tests.
>>
>> The distributed sage test output shows:
> <SNIP>
>
>> I placed test.log and the output from the upgrade in
>> sage.math.washington.edu:~justin/logs/3.4.1.rc3.
>> {test,upgrade}.log
>> respectively.
>
> Hmm, is this failure reproducible? Funny firewall settings can cause
> this and let's just say that DSage isn't exactly the most robust
> component in Sage :(

This did not reproduce. I'll look into it a bit more this weekend.
It's possible my firewall settings got in the way, but they haven't
changed since I ran this the first time. Also, there were no
unexplained or related indications of a problem in any of the system
log files.

Justin

--
Justin C. Walker, Curmudgeon-At-Large

Director
Institute for the Enhancement of the Director's Income

--------
"Weaseling out of things is what separates us from the animals.
Well, except the weasel."
- Homer J Simpson
--------