2009/12/12 Alex Ghitza <aghi...@gmail.com>:
> On Fri, Dec 11, 2009 at 10:42:15PM +0000, John Cremona wrote:
>> OK, I have just built and started testing 4.3.rc0 on my 32-bit laptop,
>> so I'll see if the same happens there, That will be with Sage's own
>> gfortran; the test failures were when I built with the system
>> fortran.
>>
>> I'll report back in the morning (GMT).
>>
Well, it's the afternoon but anyway:
On my 32-bit ubuntu laptop using Sage's own Fortran the only doctest
failures for 4.3.rc0 are the ones I also had on a different 64-bit
machine, namely:
sage -t "devel/sage/doc/en/constructions/calculus.rst"
sage -t "devel/sage/doc/en/bordeaux_2008/nf_introduction.rst"
This seems to confirm that for both of us, using a system-wide
gfortran (which is necessary to get R to build on these machines)
causes the mysterious complex conjugation effect.
What now, I wonder?
John
>> John
>>
>
> I'm curious to hear what you observe. I can indeed confirm that this
> was the problem on my end. Building Sage with SAGE_FORTRAN and
> SAGE_FORTRAN_LIB set to the system-wide gfortran avoids the error in
> building R, but results in complex(1+I) inexplicably changing the sign
> of the imaginary part. After I unset those variables (and actually
> also uninstalled the system-wide gfortran) I had to copy the system's
> libgcc_s.so to build R successfully, but then I got a properly-running
> Sage.
>
> Note that this only happens to me on 32-bit machines. On 64-bits,
> everything builds and tests fine with the system-wide gfortran.
>
> I think that until we figure out what the exact problem is (and how to
> fix it), we should hold off on
>
> http://trac.sagemath.org/sage_trac/ticket/7485
>
>
>
> Best,
> Alex
>
>
> --
> Alex Ghitza -- Lecturer in Mathematics -- The University of Melbourne
> -- Australia -- http://www.ms.unimelb.edu.au/~aghitza/
>
> --
> To post to this group, send an email to sage-devel@googlegroups.com
> To unsubscribe from this group, send an email to
> sage-devel+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/sage-devel
> URL: http://www.sagemath.org
>
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
