Memory-management crash with UTF-8 on Windows (PR#13955)
Without the file, we can do nothing with this, so please put it somewhere accessible. Also, we need exact reproduction instructions: how did you tell R this was a UTF-8 file? If you copy-pasted it, what did you copy it from? The posting guide and FAQ did ask you not to report on obsolete versions of R, so please do test a current version before replying with details of where the file is. It would also be helpful to have your exact OS, as the error appears to be in MSVCRT.dll, an operating system component.
On Thu, 17 Sep 2009, g.russell at eos-solutions.com wrote:
--=_mixed 00289247C1257634_= Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable Dear R-Bugs, thank you for your wonderful software, which we use a lot. We are having a = bit of difficulty right now because it crashes sometimes with Unicode=20 characters. Actual sessionInfo: R version 2.9.0 (2009-04-17)=20 i386-pc-mingw32=20 locale: LC=5FCOLLATE=3DGerman=5FGermany.1252;LC=5FCTYPE=3DGerman=5FGermany.1252;LC= =5FMONETARY=3DGerman=5FGermany.1252;LC=5FNUMERIC=3DC;LC=5FTIME=3DGerman=5FG= ermany.1252 attached base packages: [1] stats graphics grDevices utils datasets methods base=20 We can crash R by reading or copy-pasting the following file into it (for=20 example, Rgui --vanilla). I see a message "Microsoft Visual C++ Runtime Library", "Runtime Error!".=20 If you do not, please let me know and I'll try installing 2.9.2 and see if = that fixes the problem (the list of bug fixes doesn't seem to include this = one, but I'm not sure). Many thanks and best wishes, George Russell
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