[R] R on Windows Hangs on Packages Built from Sources

[Prof Brian Ripley]

It must be frustrating, but 

- that setup does work for other people
- I don't think we have ever seen a `hang' of that sort.

I would try one of the packages on CRAN which does have a recent binary 
compiled.  (I tend to use ash to test, as it is small and uses Fortran.)
If that fails then you need a way to debug the `hang' (and you haven't 
told us exactly where it is, so I can't suggest any).

On Wed, 5 May 2004, James Chapman wrote:

> Hi,
> 
> I am currently having the following problem with R 1.9.0. Any package that
> contains C or Fortran code hangs after I have built it and load it into R.
> I've tried this with RMySQL and udunits, originally I thought I had done
> something subtly wrong in configuring these packages; but I just had the
> same problem with a simple example R package that I wrote. I was able to
> build and load the package on a Linux system at school so I figure I've,
> hopefully, ruled out a programming problem on my part.
> 
> The system I'm having the problem on is as follows:
> AMD AthlonXP
> R 1.9.0 installed from binaries
> GCC 3.3.1

Just checking -- gcc 3.3.3 for MinGW as in

[c:/R/rw2000/src/gnuwin32]% gcc --version
gcc (GCC) 3.3.1 (mingw special 20030804-1)

?  Someone sent me several messages recently, and it transpired was trying
to use the cygwin compilers.

-- 
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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595