I've seen a few messages about incompatibilities between the Mavericks
and Snow Leopard builds of R. They have different binary build types and
packages built for one can not be used for the other.
It occurs to me that it would be helpful if R would report when a
mismatch happens between the v
- Original Message -
> From: "Richard M. Heiberger"
> To: r-sig-mac@r-project.org
> Sent: Thursday, July 17, 2014 5:03:26 AM
> Subject: [R-SIG-Mac] consistent segfault
>
> I can consistently get a segfault from Mavericks 3.1 and I suspect it
> is from a mismatch
> of compiled code. Can
I can consistently get a segfault from Mavericks 3.1 and I suspect it
is from a mismatch
of compiled code. Can this type of mismatch be detected when the
package is loaded?
When I moved to the Mavericks R distribution, I copied packages from
my 3.0 (SnowLeopard) library and then did an update.pa
For the record, if you have package tcltk loaded you can ask it which
windowing system Tk uses by
tclvalue(.Tcl("tk windowingsystem")) # "x11" or "aqua"
However, if you want to find out which it was compiled against before
loading it, you need to do the sort of thing below.
R-patched and R-