Yeah. the missing symbol is because its linking against libblas in  /
usr/lib, which  was compiled with some other probably older fortran
(why oh why can't all the fortrans get along).

I'm curious about what people think is a good solution here.
If they have a libblas in /usr/lib thats broken like this (it
requires special additional libraries but we can't deduce that at
compile time)
then as far as I know, the only way to avoid linking against their
blas  is to change the name of our blas library
i.e. libsage_blas and link against that. Or do something like set
LD_LIBRARY_PATH=$SAGE_LOCAL (but thats not a good way to solve the
problem).

Thoughts?



 
Josh

On Nov 8, 1:11 pm, mabshoff <[EMAIL PROTECTED]
dortmund.de> wrote:
> On Nov 8, 9:50 pm, Kate <[EMAIL PROTECTED]> wrote:
>
> > Michael,
>
> <SNIP>
>
> > sage -t  devel/sage-main/sage/numerical/test.py
> > **********************************************************************
> > File "test.py", line 4:
> >     : from cvxopt.base import *
> > Exception raised:
> >     Traceback (most recent call last):
> >       File "/home/kate/sage/sage-2.8.12-ia64-Linux/local/lib/python2.5/
> > doctest.py", line 1212, in __run
> >         compileflags, 1) in test.globs
> >       File "<doctest __main__.example_0[0]>", line 1, in <module>
> >         from cvxopt.base import *###line 4:
> >     : from cvxopt.base import *
> >     ImportError: /usr/lib/libblas.so.3: undefined symbol: e_wsfe
> > **********************************************************************
>
> Sorry to reply to myself so quickly: here we are pulling in the
> "local" blas and not the one we compiled with Sage. So that should be
> easily fixable.
>
> Note to self: Need more sleep.
>
> Cheers,
>
> Michael


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~----------~----~----~----~------~----~------~--~---

Reply via email to