Re: [Plplot-devel] Fwd: [Plplot-cvs] SF.net SVN: plplot:[11004] trunk/cmake/modules/python.cmake

2010-05-18 Thread Alan W. Irwin
On 2010-05-18 15:18+0100 Andrew Ross wrote: > > Werner, > > Fair point. I'll look at it later. Hi Andrew: I also thought Werner had a good idea. Your further change to implement that only made sense to me with one further small change (revision 11008) so that if the user deliberately picks Nume

Re: [Plplot-devel] Current test_diff_psc results

2010-05-18 Thread Andrew Ross
Just to update on Alan's email last week. I've now fixed up f77, f95, java and python. Tcl is also in order. The fortran changes also involved tidying up plmap / plmeridian. Example 19 use an identify transform since fortran does not support NULL functions. This does not work as the generalised c

Re: [Plplot-devel] Fwd: [Plplot-cvs] SF.net SVN: plplot:[11004] trunk/cmake/modules/python.cmake

2010-05-18 Thread Andrew Ross
Werner, Fair point. I'll look at it later. Andrew On Tue, May 18, 2010 at 10:53:15AM +0200, Werner Smekal wrote: > Hi Andrew, > > wouldn't it be better, if we disable python if numpy is not found and > print a message, that you need to set an option (WANT_NUMERIC or > similar) so that cmake te

[Plplot-devel] Fwd: [Plplot-cvs] SF.net SVN: plplot:[11004] trunk/cmake/modules/python.cmake

2010-05-18 Thread Werner Smekal
Hi Andrew, wouldn't it be better, if we disable python if numpy is not found and print a message, that you need to set an option (WANT_NUMERIC or similar) so that cmake tests for numeric and enable python if available (and skip the numpy check)? In that case people who use Numeric will definitely