There are two levels of problems 1. We can't include 64 bit compilers because we have no binaries that work for them.
2. The ar extracting hack is because we are tricking numpy's distutils into doing what we want. (More detailed explanation: Numpy's distutils has python classes for different fortran compiler types that match versions using regular expressions. These often fail on new versions. It eventually defaults to g77 which links g2c. We are tricking it into using gfortran by combining libgfortran and libf2c into libfg2c and it works (unless it can't find these). We've tried to avoid hacking the numpy distutils which automates the fortran compilation because its messy. see local/lib/python2.5/site-pacakges/numpy/distutils/fcompiler. I have an idea on how to fix this problem, we need a SAGE_f_compiler class. Josh On Jul 20, 11:30 am, Jack Schmidt <[EMAIL PROTECTED]> wrote: > Workaround for those who want to get a little further: > tar -xf sage-2.7.tar > cd sage-2.7 > mkdir -p spkg/installed > touch spkg/installed/f2c-1 > touch spkg/installed/cvxopt-0.8.2 > touch spkg/installed/gfortran-20070719 > > Now ensure that you have a symlink to g77 in your path called > "gfortran" > > make > > On Jul 20, 7:24 am, Martin Albrecht <[EMAIL PROTECTED]> > wrote: > > > Hi there, > > > the following packages failed to build/install/upgrade on my Debian/testing > > x86_64 system: > > > [gfortran] failed to recognize my gfortran installation, fixed > > spkg-installed > > already sent to William off list > > > [f2c-1] tries to copy libgfortran.a from /usr/lib while it is located in > > usr/lib/gcc/x86_64-linux-gnu/4.1/libgfortran.a. Why all the ar > > extracting/packing, can't we just -lgfortran -lf2c ? > > > On Debian/testin x86 everything upgrades fine. > > > Martin > > > -- > > name: Martin Albrecht > > _pgp:http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x8EF0DC99 > > _www:http://www.informatik.uni-bremen.de/~malb > > _jab: [EMAIL PROTECTED] --~--~---------~--~----~------------~-------~--~----~ 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/ -~----------~----~----~----~------~----~------~--~---
