Re: Problems compiling GSL on a Mac OS X

[Rodney Sparapani <rsparapa at mcw dot edu>]

> > The configure script looks for isinf().  If it doesn't find it, it can
> > get confused and tries to #define isinf(x) gsl_isinf(x).  Here is what
> > is currently in configure.in:
> >
> > AC_TRY_COMPILE([#define isinf /* */
> > #include <math.h>],[{}],[],[
> >         AC_CHECK_FUNCS(isinf, ,[
> >             AC_MSG_CHECKING([for isinf with <math.h>])
> >             AC_TRY_LINK([#include <math.h>], [float f = 0.0; 
> > isinf(f)],[
> >                 AC_MSG_RESULT(yes)
> >                 AC_DEFINE(HAVE_ISINF)],
> >                 AC_MSG_RESULT(no))])])
> >
> > --
> > Brian
I still don't understand why this is happening, but the solution is 
simple.  Just comment out the isinf and isnan macros in config.h.  
Also, if you are building from CVS you should be good to go with the 
lastest autotools since they should all be buildable and work on 
10.2.6.  I just built and used the following in a CVS build of GSL with 
shared libraries:
automake 1.7.5
autoconf 2.57
libtool 1.5
Note that I did a cvs update -d, but I never got a doc/version-ref.texi 
file so I just used the one from 1.3.  Where is this supposed to come 
from?