I think this is only a legacy question, right? In recent R, you
can/should use "parallel" instead of either multicore or snowfall.
That said, no answer if you want back compatability with older versions of R.
MW
On Mon, Jan 28, 2013 at 10:43 AM, Florian Schwaiger
<schwaige at mathematik.uni-marburg.de> wrote:> Dear R users,
>
> we have a problem when building R packages which depend on platform
> specific packages. The following example will illustrate our problem:
>
> For parallel computing (in our own package) we want to use the multicore
> package. Since multicore is not available for Windows we subtitute it by
> the snowfall package. Currently we create two packages with different
> "Depends" argument in the DESCRIPTION files (one for unix and one
for
> Windows).
>
> We considered using the .onLoad function and calling library(multicore)
> resp. library(snowfall) depending on the employed system, but it is said
> in the help of that function that it is not "Good practice" to do
so.
>
> Is there a better practice than building two packages (with the same
> name and almost same code) for the different platforms. Any help is
> appreciated.
>
>
> Matthias and Florian
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.