Re-installing PolynomF did fix that problem, though there are now others, namely R under ESS does not work. It bombs out with a curious message about text being read-only.
For me this is an equally serious problem, but that's another mailing list. You may wish to be aware of it, though. Bill. -----Original Message----- From: Prof Brian Ripley [mailto:rip...@stats.ox.ac.uk] Sent: Monday, 18 October 2010 4:20 PM To: Hodgess, Erin Cc: Duncan Murdoch; Venables, Bill (CMIS, Cleveland); r-devel@r-project.org Subject: Re: [Rd] R 2.12.0 problems with Windows XP 32-bit binary installer On Sun, 17 Oct 2010, Hodgess, Erin wrote: > I found the Rgui.exe under directory R\R-2.12.0\bin\i386. Please read the CHANGES file: this is the first item for 2.12.0. I don't believe it is related to Bill's problem. The latter looks like a package installed for R < 2.12.0 (also covered in the first item of CHANGES), but without the exact message we cannot be sure. (For the record, Mac OS X has 'fat' binaries, but R does not use them for compiled code in packages - it used subdirectories just as on Windows.) > Thanks, > Erin > > > Erin M. Hodgess, PhD > Associate Professor > Department of Computer and Mathematical Sciences > University of Houston - Downtown > mailto: hodge...@uhd.edu > > > ________________________________ > > From: Duncan Murdoch [mailto:murdoch.dun...@gmail.com] > Sent: Sun 10/17/2010 7:53 PM > To: bill.venab...@csiro.au > Cc: Hodgess, Erin; r-devel@r-project.org > Subject: Re: R 2.12.0 problems with Windows XP 32-bit binary installer > > > > bill.venab...@csiro.au wrote: >> I am having a problem with the installed R 2.12.0 on 32-bit Windows XP >> machine using the binary version. >> >> The installer I am using is R-2.12.0-win.exe . The problem I have >> is that some of my installed packages will not load, in particular >> PolynomF, which has a pre-compiled .dll. Other packages such as >> SOAR, which is pure R, attach and work without a hitch. >> >> The error message I am getting when trying to load PolynomF is a >> question whether or not the package has been compiled for the >> present architecture. (This package works without a hitch on >> R-2.11.1, for example.) >> >> Is there any significance in the name of the installer now >> including just "win" rather than "win32", as it was in the last >> release? >> >> Is it the case that all packages will have to be re-built for the >> present release? > > > Generally packages do need to be rebuilt for new versions of R (other > than patch releases, where only the last digit changes). R 2.12.0 > supports both 32 bit and 64 bit Windows in the same installation, and > since Windows doesn't have MacOS-like "universal" binaries, that means > binaries need to be put in the right place. Generally that's an i386 > subdirectory for 32 bit, and an x64 directory for 64 bit. > > I don't know why a 32 bit DLL wouldn't work with 32 bit R 2.12.0, but if > it was installed for 2.11.1, it might not be in the right place. > > Duncan Murdoch > > > > [[alternative HTML version deleted]] > > ______________________________________________ > R-devel@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel > -- Brian D. Ripley, rip...@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 ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
