[R] R 3.4.4, internet access fails on Windows XP

Duncan Murdoch murdoch.duncan at gmail.com
Thu Mar 22 10:49:47 CET 2018

On 22/03/2018 5:28 AM, Holger Taschenberger wrote:
> Dear Duncan,
>          thank you for your reply.
> On Wed, 21 Mar 2018 11:58:15 -0400
> Duncan Murdoch <murdoch.duncan at gmail.com> wrote:
>> The Windows FAQ 2.2 says, "Windows XP is no longer supported", so I think you're out of luck.  XP went past "end-of-life" in 2014.
> on <https://cran.r-project.org/bin/windows/base/README.R-3.4.4> it says
> "R 3.4.4 for Windows
> ===================
> This distribution contains a binary distribution of R-3.4.4 to run on
> Windows XP and later (including 64-bit versions of Windows) on ix86
> and x86_64 chips."

Looks as though that file wasn't updated when the FAQ was.  I am the one 
responsible for that omission, but I can't fix it now, as I retired from 
the core team last year.  The file to fix is src/gnuwin32/CRAN/ReadMe.in.

>> Other than switching to a more recent Windows version, your choices are switching to a completely different OS, or switching to an older version of R.
> R i386 runs without problems on Windows XP up to and including version 3.4.2.
> Only R i386 3.4.3 & R i386 3.4.4 cannot access the internet on my Windows XP.
> I supposed this is not so much a problem with the OS (because "IdnToAscii" exists on WindowsXP SP3 as well), but rather with the header files and/or import libraries of the MinGW environment
> Perhaps it does not take too much effort to restore WindowsXP compatibility. The described symptoms are at least easy to debug.
> I know that WindowsXP is no longer supported by Microsoft. However, I work in an academic environment and we are stuck to Windows XP on some computers because some of our data acquisition devices are only properly supported on that OS.

In that case, I'd recommend leaving R 3.4.2 on those old machines, and 
doing your main work on different ones.

> Regardless of your conclusion, I'm thankful to the R developer team to make the R software freely available to the scientific community.

Duncan Murdoch

More information about the R-help mailing list