On Jul 17, 2010, at 3:14 AM, Alan W. Irwin wrote: > On 2010-07-16 14:32-0400 Hazen Babcock wrote: > >> Alan W. Irwin wrote: >>> Set that PYTHON_HAVE_PYBUFFER macro like the usual SWIG_PYTHON macro >>> is set in bindings/python/CMakeLists.txt with the appropriate >>> CMAKE_SWIG_FLAGS list element. Of course, in that CMake logic you >>> will >>> ultimately want to avoid setting the PYTHON_HAVE_PYBUFFER macro when >>> the python version is too old to support this, but that is a nuance >>> you can deal with later if you don't have the python version >>> logic set >>> up yet. >> >> Thank you Alan, I think that I have figured out how to do this w/ >> the appropriate version checks. It seems to work on my system, but >> my version of swig is not of recent enough vintage to support >> plsmem() so I did not test the with plsmem() case. > > Thanks, Hazen, for your work on this, and thanks Simon for the > original patch. > > Simon (or Hazen) could you explain why there is a minimum swig version > requirement? Is the issue that this python capability only was > available after a certain epoch, corresponding swig adjustments had to > be made subsequent to that epoch, and those adjustments landed in the > release after 1.3.38 so our build system must test for swig version > greater than 1.3.38?
Yep. As near as I can tell the critical swig file, pybuffer.i, was not added until swig 1.3.38. I have python 2.6, but swig 1.3.36, so I can't compile with plsmem support. -Hazen ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
