[R] Compile on NT

[Guido Masarotto]

On Wed, Jun 30, 1999 at 06:57:38PM -0400, Yoon, Hoon (CICG - NY Program Trading) wrote:
> Hi,
> 
>   I am getting following error messages while trying to compile:
> 
> cp -p man/*.Rd  ../../library/base/man
> omitted
> gcc    -DHAVE_STRING_H  -c regex.c -o regex.o
> gcc: installation problem, cannot exec `cpp': No such file or directory
> make[1]: *** [regex.o] Error 1
> make: *** [rlibs] Error 2
> 
>    Is this due to the unzip32.dll? Hope not. I am trying to recompile
> packages most of all, especially modreg.
> 

  This seems a gcc installation problem. Which shell tools and compiler 
  are you using? From the message above, I am supposing you are 
  using the mingw32 version of egcs (there is no -mno-cygwin flags in the
  gcc line). Only clue, if this it true, : you get that kind of
  error ("cannot exec 'cpp'") if the environment variable
  GCC_EXEC_PREFIX is not set correctly.
  Anyway, unzip32.dll is not used before your error. So,...


> Plus, has anyone been able to extend or embed R using VC++? From what I
> hear, gcc and VC have different runtime, so this is a no go task. However,
> it is possible to use VC runtime by overwriting the original ones that comes
> with distriubution. That's why I am trying to recompile.

  If 'extend' means 'compiling packages' look to the R Venables and
  Ripley on-line complement at http://www.stats.ox.ac.uk/pub/MASS2/Compl.shtml.
  Note that full installation means more that just recompiling the dll
  and, in particular, needs more tool than the one offered by a
  standalone VC++ (perl and groff at least).

  About embedding, Brian Ripley added to the devel branch (available
  as R-devel from CRAN) some code which make this easier than in the past.
  Look for an example to the src/gnuwin32/front-ends/rtest.c file which
  shows how R.dll and the interpreter can be initialized and used.
  However, remember that R-devel should be considered experimental.
  

  I am also pretty sure that R.dll can be re-built using VC++ but this
  will require some work (and Brian additions  make this point much less
  relevant).

  guido

  
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._