I am sorry if I was not completely clear. The .a's and the .dylibs are part of the distribution with the IBM Compiler. The R CMD SHLIB works just fine.
Compiling the .f to a .o works fine. The R CMD SHLIB also works fine to generate the .so. The dyn.load of the .so fails with a set of unresolved symbols that are part of one of the libraries from IBM. So I think you answered my question. I will try the gcc route. Thanks. On May 10, 2006, at 7:03 PM, Simon Urbanek wrote: > > On May 10, 2006, at 1:02 PM, Richard Tabor wrote: > >> I am running 2.1.0 and am using the IBMXL Fortran Compiler. It has >> .a and .dylib libraries which do not seem to load with dyn.load(). >> > > How are you compiling them? I don't see how you can get either .a or > .dylib using R CMD SHLIB, so my guess is that it has nothing to do > with the compiler but with the way you compile the sources. On OS X > package libs are neither dynamic libraries (.dylib) nor static > libraries (.a) - they are bundles. The compilation shouldn't matter > and if in doubt you should be able to use gcc for the linking step in > case the IBM compiler is not compatible. > > >> Is there a way to load these libraries or convert them to .so's? >> > > There is no converting, you should get the right .so directly > otherwise you're doing something wrong. Please tell us more precisely > what you're doing. > > Cheers, > Simon _______________________________________________ R-SIG-Mac mailing list [email protected] https://stat.ethz.ch/mailman/listinfo/r-sig-mac
