[R-SIG-Mac] rgl for R-devel

Roger Koenker rkoenker at illinois.edu
Sat Jul 19 20:37:44 CEST 2014


Yes,  installing 0.94.1118  went smoothly...  although running my quantreg demo(combos) as a test crashed, but I'll assume
that this is connected to your other problem and try again later.  Thanks again to both of you

Roger

Roger Koenker
rkoenker at illinois.edu




On Jul 19, 2014, at 12:35 PM, Simon Urbanek <simon.urbanek at r-project.org> wrote:

> On Jul 19, 2014, at 10:15 AM, Roger Koenker <rkoenker at illinois.edu> wrote:
> 
>> Simon,   Maybe there is something wrong with my system, since I see:
>> 
>> yzzy: ls -l /System/Library/Frameworks/GLKit.framework
>> total 16
>> lrwxr-xr-x  1 root  wheel   22 Feb 17 23:46 GLKit -> Versions/Current/GLKit
>> lrwxr-xr-x  1 root  wheel   26 Feb 17 23:46 Resources -> Versions/Current/Resources
>> drwxr-xr-x  4 root  wheel  136 Feb 17 23:50 Versions
>> 
> 
> It's possible that your system has only the run-time and the headers are only in the SDK inside Xcode. In that case the change Duncan made should work, because the compilers are aware of the SDK and thus will find it even if it's not in /System.
> 
> Cheers,
> Simon
> 
> 
> 
>> Duncan:  I'll try the new patched version of rgl once it is available on R Forge.
>> 
>> Thanks to you both!
>> 
>> Roger
>> 
>> 
>> Thanks
>> Roger Koenker
>> rkoenker at illinois.edu
>> 
>> 
>> 
>> 
>> On Jul 18, 2014, at 8:18 PM, Simon Urbanek <simon.urbanek at r-project.org> wrote:
>> 
>>> Duncan,
>>> 
>>> I didn't look at the code, but the approach seems strange - if you use something from the GLKit framework then the include should be
>>> 
>>> #include <GLKit/GLKMathUtils.h>
>>> 
>>> without specifying -I... You should never have the need to include the "Headers" directory directly (other than cases where the framework contains non-framework headers like R does) since frameworks are searched automatically when referenced.
>>> 
>>> 
>>> @Roger: on 10.9.2 (darwin 13.1.0 - which matches your build) I see
>>> $ ls -l /System/Library/Frameworks/GLKit.framework/Headers/
>>> total 16
>>> -rw-r--r--  1 root  wheel   7327 Apr 11 05:49 GLKBaseEffect.h
>>> -rw-r--r--  1 root  wheel   1435 Apr 11 05:49 GLKEffectProperty.h
>>> [...]
>>> -rw-r--r--  1 root  wheel   1224 Apr 11 05:49 GLKMathUtils.h
>>> [...]
>>> 
>>> Cheers,
>>> Simon
>>> 
>>> 
>>> On Jul 18, 2014, at 9:07 PM, Duncan Murdoch <murdoch.duncan at gmail.com> wrote:
>>> 
>>>> On 18/07/2014, 4:41 PM, Roger Koenker wrote:
>>>>> I'm trying to
>>>>> 
>>>>> install.packages("rgl", type = "source")
>>>>> 
>>>>> with
>>>>> 
>>>>>> sessionInfo()
>>>>> R Under development (unstable) (2014-07-09 r66111)
>>>>> Platform: x86_64-apple-darwin13.1.0 (64-bit)
>>>>> 
>>>>> locale:
>>>>> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
>>>>> 
>>>>> attached base packages:
>>>>> [1] stats     graphics  grDevices utils     datasets  methods   base     
>>>>> 
>>>>> other attached packages:
>>>>> [1] quantreg_5.05 SparseM_1.05 
>>>>> 
>>>>> loaded via a namespace (and not attached):
>>>>> [1] tcltk_3.2.0 tools_3.2.0
>>>>> 
>>>>> and seeing (eventually):
>>>>> 
>>>>> clang++ -I/Library/Frameworks/R.framework/Resources/include -DNDEBUG -I/System/Library/Frameworks/GLKit.framework/Headers -I/System/Library/Frameworks/OpenGL.framework/Headers  -DHAVE_PNG_H -I/usr/local/include/libpng12 -I/usr/X11/include -DDarwin -DNO_GL_PREFIX -I/usr/X11R6/include -DHAVE_FREETYPE -Iext/ftgl -I/opt/X11/include/freetype2 -Iext -I/usr/local/include -I/usr/local/include/freetype2 -I/opt/X11/include   -Wall -mtune=core2 -g -O2  -fPIC  -Wall -mtune=core2 -g -O2  -c platform.cpp -o platform.o
>>>>> platform.cpp:16:10: fatal error: 'GLKMathUtils.h' file not found
>>>>> #include <GLKMathUtils.h> 
>>>>>      ^
>>>>> 1 error generated.
>>>>> make: *** [platform.o] Error 1
>>>>> ERROR: compilation failed for package ‘rgl’
>>>>> 
>>>>> I do have a GLKit.framework  but there is no Headers subdir ...  
>>>>> 
>>>> 
>>>> What OSX version is that?  There were warnings about deprecated code in
>>>> rgl, so I switched to using GLKMathUtils, but apparently the configure
>>>> script isn't quite right.
>>>> 
>>>> Duncan Murdoch
>>>> 
>>>> _______________________________________________
>>>> R-SIG-Mac mailing list
>>>> R-SIG-Mac at r-project.org
>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>> 



More information about the R-SIG-Mac mailing list