On Wed, May 7, 2008 at 7:32 AM, mabshoff
<[EMAIL PROTECTED]> wrote:
>
>
>
>  On May 7, 1:10 pm, "David Joyner" <[EMAIL PROTECTED]> wrote:
>  > On Wed, May 7, 2008 at 6:41 AM, mabshoff
>
>  Hi David,
>
>
>  > <[EMAIL PROTECTED]> wrote:
>  >
>  > >  On May 7, 12:29 pm, "David Joyner" <[EMAIL PROTECTED]> wrote:
>  >
>  > >  Hi David,
>  >
>  > >  > Replying to my own email again, I built 3.0.1 from for the 3rd time 
> and,
>  > >  > even, though the build goes fine apparently, at starting sage from the
>  > >  > command line, I get the same import error, even for the "uncorrupted" 
> download.
>  >
>  > >  > Any hints?
>  >
>  > >  Does the md5sum match?
>  >
>  > >  e4978d4f05528196aace3dde8b7d563b  sage-3.0.1.tar
>  >
>  > [EMAIL PROTECTED]:~/wdj/sagefiles$ md5sum sage-3.0.1.tar
>  > e4978d4f05528196aace3dde8b7d563b  sage-3.0.1.tar
>  >
>
>  ok.
>
>
>  >
>  > >  Are you still using pbuild? It would likely be the problem. If you do
>  >
>  > I thought I wasn't but
>  >
>  > [EMAIL PROTECTED]:~/wdj/sagefiles$ env | grep SAGE
>  > SAGE_BUILD_THREADS=2
>  > SAGE_PBUILD=yes
>  >
>  > so does that mean I am?
>
>  Yes. unset those variables to get rid of them.
>
>
>  > >  delete the directory
>  >
>  > >  devel/sage/build/
>  >
>  > >  Then make sure PBUILD's env variables are not defined and then run
>  > >  "sage -ba"
>  >
>  > Okay. I'm starting in a new shell, for safety's sake.
>  >
>
>  ;)
>
>  >
>  > >  Another common problem is a poisoned SAGE-env, i.e. make sure "env |
>  > >  grep SAGE" is clean. It is also not recommended to build over an
>  > >  existing build.
>  >
>  > Thanks!
>  >
>  >
>  >
>  > >  Please also post a link to install.log.
>  >
>  > It's uploading 
> tohttp://sage.math.washington.edu/home/wdj/patches/install.log
>
> > Should be there in 5-10 minutes.
>
>  Ok, I had a look and #3101 did get applied to the tree, but there is
>  some deeper issue in pbuild involved there:
>
>  While wrap.cc is now build as a C++ file the extension mwrank.so is
>  still build as a C file:
>
>  gcc -O3 -g -fwrapv -shared -fno-strict-aliasing /mnt/drive_hda1/
>  sagefiles/sage-3.0.1/devel/sage/build/temp/sage/libs/mwrank/mwrank.o -
>  L/home/wdj/wdj/sagefiles/sage-3.0.1/local/lib  -lcsage  -lcurvesntl  -
>  lg0nntl  -ljcntl  -lrankntl  -lntl  -lgmp  -lgmpxx  -lstdc++  -lm  -
>  lpari  -lstdc++  -lntl  -o /mnt/drive_hda1/sagefiles/sage-3.0.1/devel/
>  sage-main/build/sage/libs/mwrank/mwrank.so
>
>  I checked and it is supposed to be build as a C++ extension.
>
>  *But*: Checking the next extension, i.e. pow_computer_ext.so you see
>  that here gcc is also used:
>
>  gcc -O3 -g -fwrapv -shared -fno-strict-aliasing /mnt/drive_hda1/
>  sagefiles/sage-3.0.1/devel/sage/build/temp/sage/rings/padics/
>  pow_computer_ext.o -L/home/wdj/wdj/sagefiles/sage-3.0.1/local/lib  -
>  lcsage  -lcsage  -lntl  -lgmp  -lgmpxx  -lm  -lstdc++ -lstdc++  -lntl
>  -o /mnt/drive_hda1/sagefiles/sage-3.0.1/devel/sage-main/build/sage/
>  rings/padics/pow_computer_ext.so
>
>  I am not quire sure if that is supposed to be, but I always wondered
>  why on earth we would explicitly link stdc++?
>
>  Thoughts?

I have no clue. However, sage -ba works and there are no
import errors starting sage.

Thanks again!

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

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Reply via email to