[R] Re: compiling R for Windows from source

Prof Brian D Ripley ripley at stats.ox.ac.uk
Sat Apr 1 11:06:55 CEST 2000

On Fri, 31 Mar 2000, Griffith Feeney wrote:

> At 01:40 AM 3/31/00 -1000, you wrote:
> [snip]
> >> Package //L/R-1.0.0/src/library/base does not exist
> >
> >That's your problem.  Now, what do you think //L/R-1.0.0 is?  That is
> >not a Windows file specification.  Why are you using it?
> Because that's what 'tar zxvf R-1.0.0' creates. In fact, renaming it

Unlikely!  tar zxvf R-1.0.0.tgz might.

> rw1000 gives the same error message with 'rw1000' instead of 'R-1.0.0'.

You are not following carefully enough. tar can only create files. It
cannot specify a path, and something you have done has done that. It is the
"//L" which is not a Windows file specifcation.  It should be L:\R-1.0.0 or
L:/R-1.0.0.  To get anywhere you will need to find the answer to my
question.  One likely answer involves bash, as the Cygwin bash maps
L:\R-1.0.0 to //L/R-1.0.0.  

> Should src/include contain the file Rversion.h? I had to copy one from the 
> binary distribution to avoid an error message.

No, it is auto-created on a properly installed build system.  Do go back
and check that you have followed the instructions precisely: they have
worked on quite a large mumber of different systems.

One shortcut is to set RHOME and PKGD correctly in the Makefile. On your
setup I think they should be


as it is PKGD that is being quoted as an incorrect value.

But this should not be necessary, and it looks as if the shell command in
make is invoking the wrong shell (bash not sh).  You may have some
environment variable set that is not normally set, for example.

Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

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

More information about the R-help mailing list