Georg S. Weber wrote: >> c_lib will always use gcc for C code >> c_lib will use the Sun compiler for C++ if it can find it. > > It seems to me that this specific problem has been solved at least > twice in the recent past. If I remember correctly, the PolyBori > maintainers succeeded to persuade Scons to to the right things, but > never tested this myself.
Yes, they did get PolyBoRi respecting CC, so if CC is set to the Sun compiler, it will try to build with the Sun C compiler. However, PolyBoRi does so with GNU specific flags, so I do not believe the SCons/PolyBoRi issues have been resolved. See: http://sagetrac.org/sage_trac/ticket/7034 It appears one relies on SCons to chose the right flags, and quite honestly it fails to do that in all cases. > On the other hand, in the Cliquer build > system Scons was thrown out and replaced by a Makefile. Has SCons now gone from Cliquer? If so, the following tickets can probably be closed, although I would need to verify that. http://sagetrac.org/sage_trac/ticket/6852 http://sagetrac.org/sage_trac/ticket/7057 You might note my comment there "cliquer seems one seriously messed up package!" I must admit, I am yet to be convinced by SCons. It use seems to present plenty of problems in Sage. > I think the last approach would be the by far quickest one to get > "c_lib" going, since Cliquer already paved the way (and is already > made to work on OS X 10.6 / 64bit, and OS X 10.4, and many other > OS'es), so one mainly would have to carefully copy'n'paste. If William is happy with that, I'll write a makefile for c_lib. I'm no expert at writing Makefile, but I understand them more than I do SCons. > It surely would be nice to undestand how "c_lib using Scons" could be > made to work correctly in the situation you described on Sun OS, but > well, I can imagine many, many things that would be "nice to have". > > Cheers, > Georg Maybe, but perhaps just removing SCons from Sage would be even nicer! Dave --~--~---------~--~----~------------~-------~--~----~ To post to this group, send an email to [email protected] To unsubscribe from this group, send an email to [email protected] For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
