[Ifeffit] The road to Athena...

Matt Newville newville at cars.uchicago.edu
Thu Mar 7 15:34:37 CST 2002

A progress report...

== Run ifeffit from Terminal.App:

Like Chuck found, ifeffit runs (and plots) from Terminal.App once:
  setenv DISPLAY :0.0
  if( ! $?DISPLAY ) setenv DISPLAY :0.0

is in the .cshrc file and X-Darwin is running.  FWIW, putting
  setenv PGPlOT_DEV /xs

will leave the PGPLOT window up (and dormant) after quitting
ifeffit.  This dormant PGPLOT window will then be ready and
used by the ifeffit session run.  Personally, I like this
alternative, but I may use ifeffit more than others :).

== Athena:

I made some progress this morning before heading to work. I
believe I now have perl/Tk installed correctly on MacOSX,
following the instructions (and the dyld file) to the letter at
(please note the warnings about how  easy it is to trash your
OS if you get a few steps wrong!)

I couldn't run athena due to the 'other necessary modules'
(Tk::FileDialog, Chemistry::Elements,...), and so started
installing TkAtoms as I headed out the door.  Hopefully
this will auto-install all the other modules.

I didn't have loads of time and didn't try Bruce's test,
   but perl -e 'use Tk; use Ifeffit;'

returned without error for me.  I thought this was decent
success.  The 'use Ifeffit;' really _does_ call some of the
underlying Ifeffit compiled code, so I had high hopes. In
light of Chuck's problem:

>[h196132:~/Desktop] charlesb% perl test_ifeffit.pl
>dyld: perl Undefined symbols:

maybe there's more to it than I thought???  I'll look into it
-- hopefully this evening.  My guess is that some part of the
link/load stage for the perl extension isn't working correctly,
possibly because the auto-generated Makefile.PL has some
errors. I'm not sure I have any suggestions beyond going back
to the source dir and remaking the whole library:

 cd ..... src/macosx; make clean ; sudo make install
 cd ../../wrappers/perl ; perl Makefile.PL ; sudo make install

and post the output of  'perl Makefile.PL ; sudo make install'.

BR> I would guess that the swig wrapper is not compiled correctly.
BR> I barely understand swig on a nice warm fuzzy linux system, so
BR> I am not sure how much advice I have for making it work on the
BR> mac.  Some time spent at http://www.swig.org might help.  My
BR> hunch is that someone is going to have to generate a wrapper
BR> for the Mac on the Mac.

Hmmm, I'm not sure this will be necessary.  The python
extension, also using swig, worked without any changes.  With
respect to C headers and conventions, Darwin is very much BSD,
so I wouldn't blame swig just yet.


More information about the Ifeffit mailing list