[R-SIG-Mac] Mac Pro fix [Was: Error loading grDevices.so]

[Simon Urbanek]

A follow-up on this (thanks to Gene for his cooperation), a temporary  
fix is to start R using

R_LD_LIBRARY_PATH= R

a more permanent fix (with the same effect) is to uncomment/remove  
the second line in
/Library/Frameworks/R.framework/Resources/etc/i386/ldpaths

Strangely enough Mac OS X 10.4.7 (8.7.2) on Mac Pro seems to behave  
differently from other OS X variants (8.7.0=ppc, 8.7.1=CoreDuo) with  
respect to loading libraries. In general, messing with  
DYLD_LIBRARY_PATH has even more severe consequences, so it should be  
avoided at all costs.

R.app (the GUI) is not affected as it doesn't set DYLD_LIBRARY_PATH.

Cheers,
Simon


On Sep 21, 2006, at 2:28 PM, Gene wrote:

> I just tried running R from the command line for the first time with
> my new setup and get this error in trying to load grDevices.so:
>
> Error in dyn.load(x, as.logical(local), as.logical(now)) :
>    unable to load shared library '/Library/Frameworks/R.framework/
> Resources/library/grDevices/libs/ppc/grDevices.so':
> dlopen(/Library/Frameworks/R.framework/Resources/library/grDevices/
> libs/ppc/grDevices.so, 6): Symbol not found: ___eprintf
> Referenced from: /System/Library/Frameworks/CoreServices.framework/
> Versions/A/Frameworks/CFNetwork.framework/Versions/A/CFNetwork
> Expected in: /Library/Frameworks/R.framework/Resources/lib/libstdc++.
> 6.dylib
>
> I could live without quartz() or X11(), but when I get this error on
> library(stats), that really hurts.
>
> Any suggestions?  Here is my system:
>
> R.version.string =  "Version 2.3.1 (2006-06-01)"
> R GUI = Version 1.16 (3198)
> OS = 10.4.7
> Computer = 2x3GHz Mac Pro
>