error while building Sage
Mate Kosor
Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz.
I installed all prerequisites, and did
export MAKE="make -j5"
export SAGE_CHECK="yes"
make
which produced error message ending with
323 tests OK.
2 tests failed:
test_distutils test_zlib
40 tests skipped:
test_aepack test_al test_applesingle test_bsddb test_bsddb185
test_bsddb3 test_cd test_cl test_codecmaps_cn test_codecmaps_hk
test_codecmaps_jp test_codecmaps_kr test_codecmaps_tw test_curses
test_dbm test_dl test_gdbm test_gl test_imageop test_imgfile
test_kqueue test_linuxaudiodev test_macos test_macostools
test_normalization test_ossaudiodev test_pep277 test_py3kwarn
test_scriptpackages test_smtpnet test_socketserver test_ssl
test_startfile test_sunaudiodev test_timeout test_urllib2net
test_urllibnet test_winreg test_winsound test_zipfile64
4 skips unexpected on linux2:
test_dbm test_ssl test_gdbm test_bsddb
make[2]: *** [test] Greška 1
make[2]: Izlazim iz direktorija `/home/mate/sage-4.6.1/spkg/build/
python-2.6.4.p9/src'
An error occurred while testing Python
*************************************
Error testing package ** python-2.6.4.p9 **
*************************************
sage: An error occurred while testing python-2.6.4.p9
Please email sage-devel http://groups.google.com/group/sage-devel
explaining the problem and send the relevant part of
of /home/mate/sage-4.6.1/install.log. Describe your computer,
operating system, etc.
If you want to try to fix the problem yourself, *don't* just cd to
/home/mate/sage-4.6.1/spkg/build/python-2.6.4.p9 and type 'make check'
or whatever is appropriate.
Instead, the following commands setup all environment variables
correctly and load a subshell for you to debug the error:
(cd '/home/mate/sage-4.6.1/spkg/build/python-2.6.4.p9' && '/home/mate/
sage-4.6.1/sage' -sh)
When you are done debugging, you can type "exit" to leave the
subshell.
make[1]: *** [installed/python-2.6.4.p9] Greška 1
make[1]: Izlazim iz direktorija `/home/mate/sage-4.6.1/spkg'
real 13m44.906s
user 9m34.370s
sys 1m6.910s
Error building Sage.
The whole install.log is uploaded here
https://docs.google.com/leaf?id=0B-vPeW_o-3OvMzkzZmEwMzEtYTAxZi00YmIwLTlhYTUtM2E4NWNhZWE4NWMx&hl=en&authkey=CKuK240P
Jaap Spies
> I am trying to build Sage 4.6.1. on Ubuntu 10.10 notebook with
> Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz.
> I installed all prerequisites, and did
> export MAKE="make -j5"
> export SAGE_CHECK="yes"
> make
>
Don't set export SAGE_CHECK="yes"
This will fail as you noted.
Jaap
John H Palmieri
Donald Alan Morrison
> I am trying to build Sage 4.6.1. on Ubuntu 10.10 notebook with
> Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz.
> I installed all prerequisites, and did
> export MAKE="make -j5"
> export SAGE_CHECK="yes"
> make
Aside: This probably won't make any difference for you, but notice the
pre-built sage downloads are built for the LTS versions of Ubuntu
(Long Term Support -- a little more stable). If you use 10.04.1 LTS,
you can just download, extract, and run. The 8.x LTS series is end-of-
lifed.
Mate Kosor
So, I set export SAGE_CHECK="no" but the same checks went into play
and failed. Then I unset SAGE_CHECK variable and the build goes
through. I still have to do make test.
I don't know why the checks don't work for package python, but if I
could make a suggestion maybe it would be better that the check ships
that package because now the check is nonfunctional.
Also I think that SAGE_CHECK="no" should not trigger the check.
Actually, I installed Ubuntu just to be able to run Sage. So, after
what I read in the last post I went back and checked and the binary is
really entitled sage*-ubuntu_10.04.1_* I have not noticed up front
that this Ubuntu version would be prefered. Maybe a notice about this
should be written in download guide.
Thank you on the help. I'll keep you posted if something else fails
for me.
Volker Braun
Mate Kosor
David Kirkby
> I don't know why the checks don't work for package python, but if I
> could make a suggestion maybe it would be better that the check ships
> that package because now the check is nonfunctional.
IMHO that would be a bad idea. It's better to be aware of the issue
than just to cover up the failures.The actual tests that fail differ
from system to system.
It should also be noted that a large number of the packages in Sage
have no spkg-check file, so self-tests are not run, even if the
upstream source code has a test suite. I created a list of them at
some point - see the first column only of
http://trac.sagemath.org/sage_trac/ticket/9281
That list is probably not up to date. I've lost count of the number of
spkg-check files I've created, but still more need to be done.
> Also I think that SAGE_CHECK="no" should not trigger the check.
Agreed.
Five months ago John Palmieri wrote a fix for this, but as I remarked
on the ticket (#9960) endless changes (which IMHO should have been on
other tickets) were requested by the reviewer, so the patch is still
not merged. Perhaps one day it will be.
Dave
Volker Braun
On 5 February 2011 08:17, Mate Kosor <mate....@gmail.com> wrote:
> I don't know why the checks don't work for package python, but if I
> could make a suggestion maybe it would be better that the check ships
> that package because now the check is nonfunctional.IMHO that would be a bad idea. It's better to be aware of the issue
than just to cover up the failures.The actual tests that fail differ
from system to system.
David Kirkby
That's not true.
If there are not the perquisites to build a particular module, then
the tests for that module are skipped. The failures seen are actually
on parts of Python that build, but fail the tests. It seems everyone
gets a few of these.
Dave
Volker Braun
If there are not the perquisites to build a particular module, then
the tests for that module are skipped. The failures seen are actually
on parts of Python that build, but fail the tests. It seems everyone
gets a few of these.