Dear Deepayan,
Yes, that was it: My Rprofile.site file had initialization code in it
produced by the Tinn-R editor; removing this code solved the problem (which
didn't occur in R 2.7.2), as did enclosing it in if (interactive()){}.
Thank you very much for pointing me in the right direction.
John
> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
On
> Behalf Of Deepayan Sarkar
> Sent: October-21-08 7:16 PM
> To: John Fox
> Cc: r-devel@r-project.org
> Subject: Re: [Rd] R CMD INSTALL problem
>
> On Tue, Oct 21, 2008 at 3:23 PM, John Fox <[EMAIL PROTECTED]> wrote:
> > Dear list members,
> >
> > I've run into a problem with R CMD INSTALL under Windows Vista and R
2.8.0:
> >
> > --------- snip -----------
> >
> > C:\Users\John Fox\workspace>c:\R\R-2.8.0\bin\R CMD INSTALL car
> > installing to ''
>
> Not sure if this is relevant for you, but I was seeing similar
> behavior on Linux, and the reason turned out to be that the INSTALL
> script invokes
>
> R_DEFAULT_PACKAGES=NULL R --no-restore --slave
>
> which ran code in my .Rprofile, but without the usual packages
> attached. In my case, this was causing an error (because of not
> carefully written code in the profile) leading to an empty string for
> $lib (which seems to be happening for you too).
>
> -Deepayan
>
> ______________________________________________
> R-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
