openSUSE Build Service

Overview Repositories Revisions Requests Users Attributes Meta

File R-release.diff of Package R (Revision 253a5ee4fa4f3426e7abc1ba63e3d0b0)

Currently displaying revision 253a5ee4fa4f3426e7abc1ba63e3d0b0 , Show latest

# cd *inside* the ...../R-3.3.1/  directory, and do
#      zcat ..../R-release.diff.gz | patch -p1 -E
# ---------                          =====
# You need a "g[un]zip capable zcat", and canNOT use Solaris "patch"!
# Rather get ("gzip" and)  "diffutils" from any GNU mirror.
#
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/SVN-REVISION R-patched/SVN-REVISION
--- R-release-orig/SVN-REVISION	2016-06-21 09:08:01.000000000 +0200
+++ R-patched/SVN-REVISION	2016-07-29 02:43:13.300370000 +0200
@@ -1,2 +1,2 @@
-Revision: 70800
-Last Changed Date: 2016-06-21
+Revision: 70995
+Last Changed Date: 2016-07-28
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/VERSION R-patched/VERSION
--- R-release-orig/VERSION	2016-06-21 09:00:17.000000000 +0200
+++ R-patched/VERSION	2016-06-21 17:45:22.254658000 +0200
@@ -1 +1 @@
-3.3.1
+3.3.1 Patched
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/config.site R-patched/config.site
--- R-release-orig/config.site	2016-03-17 00:04:44.000000000 +0100
+++ R-patched/config.site	2016-07-10 17:45:08.852685000 +0200
@@ -256,14 +256,15 @@
 ## CXX1XFLAGS=
 ## CXX1XPICFLAGS=
 ## SHLIB_CXXIXLD=
-## SHLIB_CXXiXLDFLAGS=
+## SHLIB_CXX1XLDFLAGS=
 
 ## Additional libraries to link the internet module against.
 ## Some claim Solaris needs -lsocket -lnsl
 ## INTERNET_LIBS=
 
 ## Script from texinfo 5.1 or later.
-## Can be texi2any (the default) or makeinfo or a full path to either.
+## Usually the full path to texi2any.
+## (The name is historical: this was 'makeinfo' in texinfo 4.x)
 ## MAKEINFO=
 
 ## Full path to version of curl-config to be used
@@ -285,7 +286,8 @@
 ## OpenMP settings.
 ## The configure code is conservative about enabling OpenMP 
 ## so use can be forced here (or non-use if set but empty).
-## For packages
+## For packages (note that SHLIB_OPENMP_CXXFLAGS is for CXX, not
+## necessarily for CXX1X).
 ## SHLIB_OPENMP_CFLAGS=
 ## SHLIB_OPENMP_CXXFLAGS=
 ## SHLIB_OPENMP_FCFLAGS=
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/configure R-patched/configure
--- R-release-orig/configure	2016-06-14 00:16:32.000000000 +0200
+++ R-patched/configure	2016-07-27 02:41:07.778859749 +0200
@@ -1752,7 +1752,7 @@
                           tracemem() [no]
   --enable-R-framework[=DIR]
                           OS X only: build R framework (if possible), and
-                          specify its installation prefix [yes,
+                          specify its installation prefix [no,
                           /Library/Frameworks]
   --enable-R-shlib        build the shared/dynamic library 'libR' [no]
   --enable-R-static-lib   build the static library 'libR.a' [no]
@@ -7666,6 +7666,9 @@
   cat > conftest.c <<EOF
 int foo __attribute__ ((visibility ("hidden"))) = 1;
 int bar __attribute__ ((visibility ("default"))) = 1;
+#ifndef __GNUC__
+# error unsupported compiler
+#endif
 EOF
 r_cv_visibility_attribute=no
 if { ac_try='${CC-cc} -Werror -S conftest.c -o conftest.s 1>&5'
@@ -7723,8 +7726,8 @@
     C_VISIBILITY="-fvisibility=hidden"
   fi
 fi
-## Need to exclude Intel compilers, where this does not work.
-## The flag is documented, and is effective but also hides
+## Need to exclude Intel compilers, where this does not work correctly.
+## The flag is documented and is effective, but also hides
 ## unsatisfied references. We cannot test for GCC, as icc passes that test.
 case  "${CC}" in
   ## Intel compiler: note that -c99 may have been appended
@@ -24608,7 +24611,7 @@
   ac_cv_prog_c_openmp='none needed'
 else
   ac_cv_prog_c_openmp='unsupported'
-	  	  	  	  	  	  	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
+	  	  	  	  	            	  	            	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
                            -openmp -mp -omp -qsmp=omp -homp \
 			   -fopenmp=libomp \
                            -Popenmp --openmp; do
@@ -25057,7 +25060,7 @@
 # flags.
 r_save_CFLAGS=$CFLAGS
 CFLAGS="$CFLAGS $r_verb"
-(eval echo $as_me:25060: \"$ac_link\") >&5
+(eval echo $as_me:25063: \"$ac_link\") >&5
 r_c_v_output=`eval $ac_link 5>&1 2>&1 | grep -v 'Driving:'`
 echo "$r_c_v_output" >&5
 CFLAGS=$r_save_CFLAGS
@@ -25132,7 +25135,7 @@
 # flags.
 r_save_CFLAGS=$CFLAGS
 CFLAGS="$CFLAGS $r_cv_prog_c_v"
-(eval echo $as_me:25135: \"$ac_link\") >&5
+(eval echo $as_me:25138: \"$ac_link\") >&5
 r_c_v_output=`eval $ac_link 5>&1 2>&1 | grep -v 'Driving:'`
 echo "$r_c_v_output" >&5
 CFLAGS=$r_save_CFLAGS
@@ -26113,7 +26116,7 @@
   ac_cv_prog_f77_openmp='none needed'
 else
   ac_cv_prog_f77_openmp='unsupported'
-	  	  	  	  	  	  	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
+	  	  	  	  	            	  	            	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
                            -openmp -mp -omp -qsmp=omp -homp \
 			   -fopenmp=libomp \
                            -Popenmp --openmp; do
@@ -26244,7 +26247,7 @@
   ac_cv_prog_cxx_openmp='none needed'
 else
   ac_cv_prog_cxx_openmp='unsupported'
-	  	  	  	  	  	  	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
+	  	  	  	  	            	  	            	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
                            -openmp -mp -omp -qsmp=omp -homp \
 			   -fopenmp=libomp \
                            -Popenmp --openmp; do
@@ -27473,6 +27476,7 @@
     fi
     ;;
   darwin*)
+    ## which these days mean macOS
     $as_echo "#define HAVE_NO_SYMBOL_UNDERSCORE 1" >>confdefs.h
 
     ;;
@@ -27544,6 +27548,7 @@
 
     ;;
   irix*)
+    ## final release in 2006
     { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler needs -OPT:IEEE_NaN_inf=ON" >&5
 $as_echo_n "checking whether C compiler needs -OPT:IEEE_NaN_inf=ON... " >&6; }
 if ${r_cv_c_optieee+:} false; then :
@@ -27806,6 +27811,7 @@
     fi
     ;;
   osf*)
+    ## does that really still exist?
     $as_echo "#define HAVE_NO_SYMBOL_UNDERSCORE 1" >>confdefs.h
 
     ;;
@@ -35695,8 +35701,8 @@
 { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether bzip2 support suffices" >&5
 $as_echo_n "checking whether bzip2 support suffices... " >&6; }
 if test "x${have_bzlib}" = xyes; then
-  { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
-$as_echo "no" >&6; }
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
   LIBS="-lbz2 ${LIBS}"
 else
   as_fn_error $? "bzip2 library and headers are required" "$LINENO" 5
@@ -36161,7 +36167,7 @@
     CURL_LIBS=`${CURL_CONFIG} --libs`
   fi
 fi
-r_save_CPPFLAGS="${CPPLAGS}"
+r_save_CPPFLAGS="${CPPFLAGS}"
 CPPFLAGS="${CURL_CPPFLAGS} ${CPPFLAGS}"
 r_save_LIBS="${LIBS}"
 LIBS="${CURL_LIBS} ${LIBS}"
@@ -41310,7 +41316,7 @@
   ac_cv_prog_fc_openmp='none needed'
 else
   ac_cv_prog_fc_openmp='unsupported'
-	  	  	  	  	  	  	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
+	  	  	  	  	            	  	            	  	  	  	                                	  	  	  	  	  	  for ac_option in -fopenmp -xopenmp -qopenmp \
                            -openmp -mp -omp -qsmp=omp -homp \
 			   -fopenmp=libomp \
                            -Popenmp --openmp; do
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/configure.ac R-patched/configure.ac
--- R-release-orig/configure.ac	2016-03-17 00:04:40.000000000 +0100
+++ R-patched/configure.ac	2016-07-24 17:45:11.015809000 +0200
@@ -222,7 +222,7 @@
 ## Allow the user to specify building an R framework (Darwin).
 AC_ARG_ENABLE([R-framework],
 [AS_HELP_STRING([--enable-R-framework@<:@=DIR@:>@],[OS X only: build R framework (if possible), and specify
-                 its installation prefix @<:@yes, /Library/Frameworks@:>@])],
+                 its installation prefix @<:@no, /Library/Frameworks@:>@])],
 [want_R_framework="${enableval}"],
 [want_R_framework=no])
 ## Can only build frameworks on Darwin.
@@ -1004,6 +1004,7 @@
     fi
     ;;
   darwin*)
+    ## which these days mean macOS
     AC_DEFINE(HAVE_NO_SYMBOL_UNDERSCORE)
     ;;
   hpux*)
@@ -1020,6 +1021,7 @@
     AC_CHECK_LIB(dld, shl_load, [R_XTRA_LIBS="-ldld ${R_XTRA_LIBS}"])
     ;;
   irix*)
+    ## final release in 2006
     R_C_OPTIEEE
     ## <NOTE>
     ## We really should test explictly whether the Fortran and C++
@@ -1128,6 +1130,7 @@
     fi
     ;;
   osf*)
+    ## does that really still exist?
     AC_DEFINE(HAVE_NO_SYMBOL_UNDERSCORE)
     ;;  
 esac
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/COPYRIGHTS R-patched/doc/COPYRIGHTS
--- R-release-orig/doc/COPYRIGHTS	2016-03-17 00:04:41.000000000 +0100
+++ R-patched/doc/COPYRIGHTS	2016-07-12 02:41:42.042708000 +0200
@@ -10,7 +10,8 @@
 3: most source files contain a copyright statement allowing use of
 that file under GPL version 2 or later: the main exceptions are the
 included versions of packages 'MASS', 'class', 'nnet', 'rpart' and
-'spatial' (GPL-2 or GPL-3).
+'spatial' (GPL-2 or GPL-3).  (The auxiliary file m4/openmp.m4 is under
+GPL-3, but its incorporation into 'configure' is not.)
 
 The status of files used only in the Windows port is in file
 src/gnuwin32/COPYRIGHTS.win, which is appended to this file in binary
@@ -369,30 +370,33 @@
 For binary builds of R that requires us to include
 
   Redistribution and use in source and binary forms, with or without
-  modification, are permitted provided that the following conditions are met:
+  modification, are permitted provided that the following conditions
+  are met:
 
-      * Redistributions of source code must retain the above copyright notice,
-	this list of conditions and the following disclaimer.
+      * Redistributions of source code must retain the above copyright
+	notice, this list of conditions and the following disclaimer.
 
-      * Redistributions in binary form must reproduce the above copyright
-	notice, this list of conditions and the following disclaimer in the
-	documentation and/or other materials provided with the distribution.
-
-      * Neither the name of the University of Cambridge nor the name of Google
-	Inc. nor the names of their contributors may be used to endorse or
-	promote products derived from this software without specific prior
-	written permission.
-
-  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
-  AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
-  IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
-  ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
-  LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
-  CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
-  SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
-  INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
-  CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
-  ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+      * Redistributions in binary form must reproduce the above
+	copyright notice, this list of conditions and the following
+	disclaimer in the documentation and/or other materials
+	provided with the distribution.
+
+      * Neither the name of the University of Cambridge nor the name
+	of Google Inc. nor the names of their contributors may be used
+	to endorse or promote products derived from this software
+	without specific prior written permission.
+
+  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+  FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+  COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+  INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
+  BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+  LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+  CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+  LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
+  ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
   POSSIBILITY OF SUCH DAMAGE.
 
 
@@ -514,6 +518,12 @@
 
    Copyright various dates Free Software Foundation
 
+m4/openmp.m4
+  Copyright (C) 2001-2012 Free Software Foundation, Inc.
+  Copyright (C) 2015-2016 R Core Team
+
+  Under GPL-3 with the Autoconf Configure Script Exception, version 3.0.
+
 tools/config.quess
 tools/config.sub
 
@@ -583,3 +593,80 @@
 
    The design of the current logo is based on the previous logo that
    was included in the R source from 1998 to 2016.
+
+	---------------------------------------------------
+
+Binary distributions of R may include compiled code from the PCRE
+library whose licence is:
+
+  THE BASIC LIBRARY FUNCTIONS
+  ---------------------------
+  Written by:       Philip Hazel
+  Email local part: ph10
+  Email domain:     cam.ac.uk
+
+  University of Cambridge Computing Service, Cambridge, England.
+
+  Copyright (c) 1997-2015 University of Cambridge
+  All rights reserved.
+
+  PCRE JUST-IN-TIME COMPILATION SUPPORT
+  -------------------------------------
+  Written by:       Zoltan Herczeg
+  Email local part: hzmester
+  Emain domain:     freemail.hu
+
+  Copyright(c) 2010-2015 Zoltan Herczeg
+  All rights reserved.
+
+  STACK-LESS JUST-IN-TIME COMPILER
+  --------------------------------
+  Written by:       Zoltan Herczeg
+  Email local part: hzmester
+  Emain domain:     freemail.hu
+
+  Copyright(c) 2009-2015 Zoltan Herczeg
+  All rights reserved.
+
+  THE C++ WRAPPER FUNCTIONS
+  -------------------------
+  Contributed by:   Google Inc.
+
+  Copyright (c) 2007-2012, Google Inc.
+  All rights reserved.
+
+
+  THE "BSD" LICENCE
+  -----------------
+
+  Redistribution and use in source and binary forms, with or without
+  modification, are permitted provided that the following conditions
+  are met:
+
+      * Redistributions of source code must retain the above copyright
+        notice, this list of conditions and the following disclaimer.
+
+      * Redistributions in binary form must reproduce the above
+        copyright notice, this list of conditions and the following
+        disclaimer in the documentation and/or other materials
+        provided with the distribution.
+
+      * Neither the name of the University of Cambridge nor the name
+        of Google Inc. nor the names of their contributors may be used
+        to endorse or promote products derived from this software
+        without specific prior written permission.
+
+  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+  FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+  COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+  INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
+  BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+  LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+  CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+  LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
+  ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+  POSSIBILITY OF SUCH DAMAGE.
+
+	---------------------------------------------------
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/FAQ R-patched/doc/FAQ
--- R-release-orig/doc/FAQ	2016-06-21 09:08:01.000000000 +0200
+++ R-patched/doc/FAQ	2016-07-29 02:43:12.014904000 +0200
@@ -246,11 +246,12 @@
 
    The current version of R will configure and build under a number of
 common Unix-like (e.g., <https://en.wikipedia.org/wiki/Unix-like>)
-platforms including CPU-linux-gnu for the i386, amd64, alpha, arm/armel,
-hppa, ia64, m68k, mips/mipsel, powerpc, s390 and sparc CPUs (e.g.,
+platforms including CPU-linux-gnu for the i386, amd64/x86_64, alpha,
+arm, arm64, hppa, mips/mipsel, powerpc, s390x and sparc CPUs (e.g.,
 <https://buildd.debian.org/build.php?&pkg=r-base>), i386-hurd-gnu,
-CPU-kfreebsd-gnu for i386 and amd64, powerpc-apple-darwin,
-mips-sgi-irix, i386-freebsd, rs6000-ibm-aix, and sparc-sun-solaris.
+CPU-kfreebsd-gnu for i386 and amd64, i386-pc-solaris, rs6000-ibm-aix,
+sparc-sun-solaris, x86_64-apple-darwin, x86_64-unknown-freebsd and
+x86_64-unknown-openbsd.
 
    If you know about other platforms, please drop us a note.
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/NEWS R-patched/doc/NEWS
--- R-release-orig/doc/NEWS	2016-06-21 09:13:46.000000000 +0200
+++ R-patched/doc/NEWS	2016-07-29 02:54:47.585891000 +0200
@@ -1,5 +1,48 @@
 R News
 
+CHANGES IN R 3.3.1 patched:
+
+  NEW FEATURES:
+
+    * extSoftVersion() now reports the version (if any) of the readline
+      library in use.
+
+    * Convenience function hasName() has been added; it is intended to
+      replace the common idiom !is.null(x$name) without the usually
+      unintended partial name matching.
+
+  INSTALLATION and INCLUDED SOFTWARE:
+
+    * Versions of the readline library >= 6.3 had been changed so that
+      terminal window resizes were not signalled to readline: code has
+      been added using a explicit signal handler to work around that
+      (when R is compiled against readline >= 6.3).  (PR#16604)
+
+  BUG FIXES:
+
+    * The check for non-portable flags in R CMD check could be stymied
+      by src/Makevars files which contained targets.
+
+    * (Windows only) When using certain desktop themes in Windows 7 or
+      higher, Alt-Tab could cause Rterm to stop accepting input.
+      (PR#14406; patch submitted by Jan Gleixner.)
+
+    * pretty(d, ..) behaves better for date-time d (PR#16923).
+
+    * When a class name matches multiple classes in the cache, perform
+      a dynamic search in order to obey namespace imports.  This should
+      eliminate annoying messages about multiple hits in the class
+      cache.  Also, pass along the package from the ClassExtends object
+      when looking up superclasses in the cache.
+
+    * sample(NA_real_) now works.
+
+    * Packages using non-ASCII encodings in their code did not install
+      data properly on systems using different encodings.
+
+    * merge(df1, df2) now also works for data frames with column names
+      "na.last", "decreasing", or "method".  (PR#17119)
+
 CHANGES IN R 3.3.1:
 
   BUG FIXES:
@@ -22,7 +65,7 @@
 
     * pretty(d, ..) for date-time d rarely failed when "halfmonth" time
       steps were tried (PR#16923) and on 'inaccurate' platforms such as
-      32-bit windows or a configuration with --disable-long-double; see
+      32-bit Windows or a configuration with --disable-long-double; see
       comment #15 of PR#16761.
 
     * In text.default(x, y, labels), the rarely(?) used default for
@@ -35,7 +78,8 @@
 
     * Use of Ctrl-C to terminate a reverse incremental search started
       by Ctrl-R in the readline-based Unix terminal interface is now
-      supported for readline >= 6.3 (Ctrl-G always worked).  (PR#16603)
+      supported when R was compiled against readline >= 6.0 (Ctrl-G
+      always worked).  (PR#16603)
 
     * diff(<difftime>) now keeps the "units" attribute, as subtraction
       already did, PR#16940.
@@ -852,7 +896,7 @@
       handle UTF-8 strings properly.  (PR#16264)
 
     * read.dcf() sometimes misread lines longer than 8191 characters.
-      (Reported by Herv'e Pag`es with a patch.)
+      (Reported by Herv? Pag?s with a patch.)
 
     * within(df, ..) no longer drops columns whose name start with a
       ".".
@@ -869,7 +913,7 @@
       with non-small sizes.
 
     * If the expression in the traceback was too long, traceback() did
-      not report the source line number.  (Patch by Kirill M"uller.)
+      not report the source line number.  (Patch by Kirill M?ller.)
 
     * The browser did not truncate the display of the function when
       exiting with options("deparse.max.lines") set.  (PR#16581)
@@ -1957,7 +2001,7 @@
     * R CMD check now by default checks code usage directly on the
       package namespace without loading and attaching the package and
       its suggests and enhances.  For good practice with packages in
-      the Suggests field, see SS1.1.3.1 of 'Writing R Extensions'.  For
+      the Suggests field, see ?1.1.3.1 of 'Writing R Extensions'.  For
       use of lazy-data objects in the package's own code, see ?data.
 
   BUG FIXES:
@@ -3562,8 +3606,8 @@
       --encoding=foo were not handled properly.  (PR#15405)
 
     * If x is a symbol, is.vector(x, "name") now returns TRUE, since
-      "name" and "symbol" should be synonyms.  (Reported by Herv'e
-      Pag`es.)
+      "name" and "symbol" should be synonyms.  (Reported by Herv?
+      Pag?s.)
 
     * R CMD rtags works on platforms (such as OS X) with a
       XSI-conformant shell command echo. (PR#15231)
@@ -3729,7 +3773,7 @@
 
     * as.dendrogram()'s hclust method uses less memory and hence gets
       considerably faster for large (n ~ 1000) clusterings, thanks to
-      Daniel M"ullner. (PR#15174)
+      Daniel M?llner. (PR#15174)
 
     * The return value when all workers failed from
       parallel::mclapply(mc.preschedule = TRUE) was a list of strings
@@ -4036,7 +4080,7 @@
 
     * mapply(FUN, a1, ..) now also works when a1 (or a further such
       argument) needs a length() method (which the documented arguments
-      never do).  (Requested by Herv'e Pag`es; with a patch.)
+      never do).  (Requested by Herv? Pag?s; with a patch.)
 
     * .onDetach() is supported as an alternative to .Last.lib.  Unlike
       .Last.lib, this does not need to be exported from the package's
@@ -4586,7 +4630,7 @@
       so can be omitted if desired.
 
     * The markup for many messages has been changed to make them easier
-      to translate, incorporating suggestions from Lukasz Daniel.
+      to translate, incorporating suggestions from ?ukasz Daniel.
 
   INSTALLATION:
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/NEWS.Rd R-patched/doc/NEWS.Rd
--- R-release-orig/doc/NEWS.Rd	2016-06-12 00:15:12.000000000 +0200
+++ R-patched/doc/NEWS.Rd	2016-07-24 02:41:26.223307000 +0200
@@ -5,6 +5,60 @@
 \title{R News}
 \encoding{UTF-8}
 
+\section{\Rlogo CHANGES IN R 3.3.1 patched}{
+  \subsection{NEW FEATURES}{
+    \itemize{
+      \item \code{extSoftVersion()} now reports the version (if any) of
+      the \code{readline} library in use.
+
+      \item Convenience function \code{hasName()} has been added; it
+      is intended to replace the common idiom \code{!is.null(x$name)}
+      without the usually unintended partial name matching.
+    }
+  }
+
+  \subsection{INSTALLATION and INCLUDED SOFTWARE}{
+    \itemize{
+      \item Versions of the \code{readline} library >= 6.3 had been
+      changed so that terminal window resizes were not signalled to
+      \code{readline}: code has been added using a explicit signal
+      handler to work around that (when \R is compiled against readline
+      >= 6.3).  (\PR{16604})
+    }
+  }
+
+  \subsection{BUG FIXES}{
+    \itemize{
+      \item The check for non-portable flags in \command{R CMD check}
+      could be stymied by \file{src/Makevars} files which contained
+      targets.
+
+      \item (Windows only) When using certain desktop themes in Windows
+      7 or higher, \kbd{Alt-Tab} could cause \command{Rterm} to stop
+      accepting input.  (\PR{14406}; patch submitted by Jan Gleixner.)
+
+      \item \code{pretty(d, ..)} behaves better for date-time \code{d}
+      (\PR{16923}).
+
+      \item When a class name matches multiple classes in the cache,
+      perform a dynamic search in order to obey namespace imports.  This
+      should eliminate annoying messages about multiple hits in the
+      class cache.  Also, pass along the package from the
+      \code{ClassExtends} object when looking up superclasses in the
+      cache.
+
+      \item \code{sample(NA_real_)} now works.
+      
+      \item Packages using non-ASCII encodings in their code did not 
+      install data properly on systems using different encodings.
+
+      \item \code{merge(df1, df2)} now also works for data frames with
+      column names \code{"na.last"}, \code{"decreasing"}, or
+      \code{"method"}.  (\PR{17119})
+    }
+  }
+}
+
 \section{\Rlogo CHANGES IN R 3.3.1}{
   \subsection{BUG FIXES}{
     \itemize{
@@ -28,7 +82,7 @@
 
       \item \code{pretty(d, ..)} for date-time \code{d} rarely failed
       when \code{"halfmonth"} time steps were tried (\PR{16923}) and
-      on \sQuote{inaccurate} platforms such as 32-bit windows or
+      on \sQuote{inaccurate} platforms such as 32-bit Windows or
       a configuration with \command{--disable-long-double}; see comment
       #15 of \PR{16761}.
 
@@ -43,8 +97,9 @@
 
       \item Use of \code{Ctrl-C} to terminate a reverse incremental
       search started by \code{Ctrl-R} in the \code{readline}-based Unix
-      terminal interface is now supported for \code{readline} >= 6.3
-      (\code{Ctrl-G} always worked).  (\PR{16603})
+      terminal interface is now supported when \R{} was compiled against
+      \code{readline} >= 6.0 (\code{Ctrl-G} always worked).
+      (\PR{16603})
 
       \item \code{diff(<difftime>)} now keeps the \code{"units"}
       attribute, as subtraction already did, \PR{16940}.
Binary files R-release-orig/doc/NEWS.pdf and R-patched/doc/NEWS.pdf differ
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/html/NEWS.html R-patched/doc/html/NEWS.html
--- R-release-orig/doc/html/NEWS.html	2016-06-21 09:13:46.000000000 +0200
+++ R-patched/doc/html/NEWS.html	2016-07-29 02:54:50.341714235 +0200
@@ -7,6 +7,84 @@
 
 <h2>R News</h2>
 
+<h3><img src="../help/figures/../../html/Rlogo.svg" class="toplogo" alt="[R logo]" /> CHANGES IN R 3.3.1 patched</h3>
+
+
+
+<h4>NEW FEATURES</h4>
+
+
+<ul>
+<li> <p><code>extSoftVersion()</code> now reports the version (if any) of
+the <code>readline</code> library in use.
+</p>
+</li>
+<li><p> Convenience function <code>hasName()</code> has been added; it
+is intended to replace the common idiom <code>!is.null(x$name)</code>
+without the usually unintended partial name matching.
+</p>
+</li></ul>
+
+
+
+
+<h4>INSTALLATION and INCLUDED SOFTWARE</h4>
+
+
+<ul>
+<li><p> Versions of the <code>readline</code> library &gt;= 6.3 had been
+changed so that terminal window resizes were not signalled to
+<code>readline</code>: code has been added using a explicit signal
+handler to work around that (when <span style="font-family: Courier New, Courier; color: #666666;"><b>R</b></span> is compiled against readline
+&gt;= 6.3).  (<a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=16604">PR#16604</a>)
+</p>
+</li></ul>
+
+
+
+
+<h4>BUG FIXES</h4>
+
+
+<ul>
+<li><p> The check for non-portable flags in <code>R CMD check</code>
+could be stymied by &lsquo;<span class="file">src/Makevars</span>&rsquo; files which contained
+targets.
+</p>
+</li>
+<li><p> (Windows only) When using certain desktop themes in Windows
+7 or higher, <kbd>Alt-Tab</kbd> could cause <code>Rterm</code> to stop
+accepting input.  (<a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=14406">PR#14406</a>; patch submitted by Jan Gleixner.)
+</p>
+</li>
+<li> <p><code>pretty(d, ..)</code> behaves better for date-time <code>d</code>
+(<a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=16923">PR#16923</a>).
+</p>
+</li>
+<li><p> When a class name matches multiple classes in the cache,
+perform a dynamic search in order to obey namespace imports.  This
+should eliminate annoying messages about multiple hits in the
+class cache.  Also, pass along the package from the
+<code>ClassExtends</code> object when looking up superclasses in the
+cache.
+</p>
+</li>
+<li> <p><code>sample(NA_real_)</code> now works.
+</p>
+</li>
+<li><p> Packages using non-ASCII encodings in their code did not 
+install data properly on systems using different encodings.
+</p>
+</li>
+<li> <p><code>merge(df1, df2)</code> now also works for data frames with
+column names <code>"na.last"</code>, <code>"decreasing"</code>, or
+<code>"method"</code>.  (<a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=17119">PR#17119</a>)
+</p>
+</li></ul>
+
+
+
+
 <h3><img src="../help/figures/../../html/Rlogo.svg" class="toplogo" alt="[R logo]" /> CHANGES IN R 3.3.1</h3>
 
 
@@ -40,7 +118,7 @@
 </li>
 <li> <p><code>pretty(d, ..)</code> for date-time <code>d</code> rarely failed
 when <code>"halfmonth"</code> time steps were tried (<a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=16923">PR#16923</a>) and
-on &lsquo;inaccurate&rsquo; platforms such as 32-bit windows or
+on &lsquo;inaccurate&rsquo; platforms such as 32-bit Windows or
 a configuration with <code>--disable-long-double</code>; see comment
 #15 of <a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=16761">PR#16761</a>.
 </p>
@@ -59,8 +137,9 @@
 </li>
 <li><p> Use of <code>Ctrl-C</code> to terminate a reverse incremental
 search started by <code>Ctrl-R</code> in the <code>readline</code>-based Unix
-terminal interface is now supported for <code>readline</code> &gt;= 6.3
-(<code>Ctrl-G</code> always worked).  (<a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=16603">PR#16603</a>)
+terminal interface is now supported when <span style="font-family: Courier New, Courier; color: #666666;"><b>R</b></span> was compiled against
+<code>readline</code> &gt;= 6.0 (<code>Ctrl-G</code> always worked).
+(<a href="https://bugs.R-project.org/bugzilla3/show_bug.cgi?id=16603">PR#16603</a>)
 </p>
 </li>
 <li> <p><code>diff(&lt;difftime&gt;)</code> now keeps the <code>"units"</code>
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/html/R-admin.html R-patched/doc/html/R-admin.html
--- R-release-orig/doc/html/R-admin.html	2016-06-21 09:20:40.000000000 +0200
+++ R-patched/doc/html/R-admin.html	2016-07-29 03:05:29.838612295 +0200
@@ -1,6 +1,6 @@
 <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
 <html>
-<!-- This manual is for R, version 3.3.1 (2016-06-21).
+<!-- This manual is for R, version 3.3.1 Patched (2016-07-28).
 
 Copyright (C) 2001-2016 R Core Team
 
@@ -183,6 +183,7 @@
     <li><a name="toc-Other-Options-1" href="#Other-Options">2.7 Other Options</a>
     <ul class="no-bullet">
       <li><a name="toc-OpenMP-Support-1" href="#OpenMP-Support">2.7.1 OpenMP Support</a></li>
+      <li><a name="toc-C_002b_002b-Support-1" href="#C_002b_002b-Support">2.7.2 C++ Support</a></li>
     </ul></li>
     <li><a name="toc-Testing-an-Installation" href="#Testing-a-Unix_002dalike-Installation">2.8 Testing an Installation</a></li>
   </ul></li>
@@ -292,7 +293,6 @@
     <ul class="no-bullet">
       <li><a name="toc-Clang-1" href="#Clang">C.2.1 Clang</a></li>
       <li><a name="toc-Intel-compilers-1" href="#Intel-compilers">C.2.2 Intel compilers</a></li>
-      <li><a name="toc-Oracle-Solaris-Studio-compilers-1" href="#Oracle-Solaris-Studio-compilers">C.2.3 Oracle Solaris Studio compilers</a></li>
     </ul></li>
     <li><a name="toc-OS-X-2" href="#OS-X">C.3 OS X</a>
     <ul class="no-bullet">
@@ -333,7 +333,7 @@
 
 <p>This is a guide to installation and administration for R.
 </p>
-<p>This manual is for R, version 3.3.1 (2016-06-21).
+<p>This manual is for R, version 3.3.1 Patched (2016-07-28).
 </p>
 <p>Copyright &copy; 2001&ndash;2016 R Core Team
 </p>
@@ -1216,6 +1216,59 @@
 that overridden values are not tested.
 </p>
 <hr>
+<a name="C_002b_002b-Support"></a>
+<a name="C_002b_002b-Support-1"></a>
+<h4 class="subsection">2.7.2 C++ Support</h4>
+
+<p>C++ is not used by R itself, but support is provided for installing
+packages with C++ code via <code>make</code> macros defined in file
+<samp>etc/Makeconf</samp> (and with explanations in file <samp>config.site</samp>):
+</p><div class="example">
+<pre class="example">CXX
+CXXFLAGS
+CXXPICFLAGS
+
+CXX1X
+CXX1XSTD
+CXX1XFLAGS
+CXX1XPICFLAGS
+</pre></div>
+<p>The macros <code>CXX</code> etc are those used by default for C++ code: their
+values are not checked.  <code>configure</code> will attempt to set the
+rest suitably, choosing for <code>CXX1XSTD</code> a suitable flag such as
+<samp>-std=c++11</samp> for C++11 support.  The inferred values can be
+overridden in file <samp>config.site</samp> or on the <code>configure</code>
+command line: user-supplied values will be tested compiling some C++11
+code.
+</p>
+<p>It may be<a name="DOCF12" href="#FOOT12"><sup>12</sup></a> that there is no suitable flag for C++11 support, in which
+case a different compiler could be selected for <code>CXX1X</code> and its
+corresponding flags.  Some compilers<a name="DOCF13" href="#FOOT13"><sup>13</sup></a> by default assume a later
+standard than C++98 whereas the latter is assumed by some packages.  So
+users of GCC 6 might like to specify
+</p><div class="example">
+<pre class="example">CXX='g++ -std=gnu++98'
+CXX1X=g++
+CXX1XSTD='-std=c++11'
+</pre></div>
+
+<p>The <samp>-std</samp> flag is supported by the GCC, <code>clang</code>, Intel
+and Solaris compilers (the latter from version 12.4).  Currently
+accepted values are (plus some synonyms)
+</p><div class="example">
+<pre class="example">g++:     c++98 gnu++98 c++11 gnu+11 c++14 gnu++14 c++1z gnu++1z
+Intel:   gnu+98 c++11 c++14 (from 16.0)
+Solaris: c++03 c++11 c++14 (from 12.5)
+</pre></div>
+<p>(Those for <code>clang++</code> are not documented, but seem to be based on
+<code>g++</code>.)  Versions 4.3.x to 4.8.x of <code>g++</code> accepted flag
+<samp>-std=c++0x</samp> with partial support<a name="DOCF14" href="#FOOT14"><sup>14</sup></a> for C++11:
+this is currently still accepted as a deprecated synonym for
+<samp>-std=c++11</samp>.
+</p>
+<p>For the use of C++11 in R packages see the &lsquo;Writing R Extensions&rsquo; manual.
+</p>
+<hr>
 <a name="Testing-a-Unix_002dalike-Installation"></a>
 <a name="Testing-an-Installation"></a>
 <h3 class="section">2.8 Testing an Installation</h3>
@@ -1321,7 +1374,7 @@
 <em>must</em> be installed in a path that does not contain spaces.
 </p>
 <p>Installation is <em>via</em> the installer
-<samp>R-3.3.1-win.exe</samp>.  Just double-click on the icon and
+<samp>R-3.3.1patched-win.exe</samp>.  Just double-click on the icon and
 follow the instructions.  When installing on a 64-bit version of Windows
 the options will include 32- or 64-bit versions of R (and the default is
 to install both).  You can uninstall R from the Control Panel.
@@ -1568,7 +1621,7 @@
 <a name="Support-for-libcurl-1"></a>
 <h4 class="subsection">3.1.6 Support for libcurl</h4>
 
-<p><code>libcurl</code> version 7.28.0 or later can be used to support
+<p><code>libcurl</code> version 7.28.0 or later is used to support
 <code>curlGetHeaders</code> and the <code>&quot;libcurl&quot;</code> methods of
 <code>download.file</code> and <code>url</code>.
 </p>
@@ -1652,7 +1705,7 @@
 <samp>MkRules.local</samp> and edit it.)  The <strong>texinfo</strong> 5.x package for
 use on Windows is available at
 <a href="https://www.stats.ox.ac.uk/pub/Rtools/">https://www.stats.ox.ac.uk/pub/Rtools/</a>: you will also need to
-install <code>Perl</code><a name="DOCF12" href="#FOOT12"><sup>12</sup></a>
+install <code>Perl</code><a name="DOCF15" href="#FOOT15"><sup>15</sup></a>
 </p>
 <hr>
 <a name="Building-the-Inno-Setup-installer"></a>
@@ -1719,7 +1772,7 @@
 installation (in double quotes if it contains spaces or backslashes).
 </p>
 <p>Both methods create an executable with a standard name such as
-<samp>R-3.3.1-win.exe</samp>, so please rename it to indicate that
+<samp>R-3.3.1patched-win.exe</samp>, so please rename it to indicate that
 it is customized.  If you intend to <em>distribute</em> a customized
 installer please do check that license requirements are met &ndash; note that
 the installer will state that the contents are distributed under GPL
@@ -1778,21 +1831,21 @@
 </pre></div>
 
 <p>which will result in a file with a name like
-<samp>R-3.3.1-win32.msi</samp>.  This can be double-clicked to be
+<samp>R-3.3.1patched-win32.msi</samp>.  This can be double-clicked to be
 installed, but those who need it will know what to do with it (usually
 by running <code>msiexec /i</code> with additional options).  Properties
 that users might want to set from the <code>msiexec</code> command line
 include &lsquo;<samp>ALLUSERS</samp>&rsquo;, &lsquo;<samp>INSTALLDIR</samp>&rsquo; (something like
-<samp>c:\Program Files\R\R-3.3.1</samp>) and &lsquo;<samp>RMENU</samp>&rsquo; (the path
+<samp>c:\Program Files\R\R-3.3.1patched</samp>) and &lsquo;<samp>RMENU</samp>&rsquo; (the path
 to the &lsquo;<samp>R</samp>&rsquo; folder on the start menu) and &lsquo;<samp>STARTDIR</samp>&rsquo; (the
 starting directory for R shortcuts, defaulting to something like
 <samp>c:\Users\name\Documents\R</samp>).
 </p>
 <p>The MSI installer can be built both from a 32-bit build of R
-(<samp>R-3.3.1-win32.msi</samp>) and from a 64-bit build of R
-(<samp>R-3.3.1-win64.msi</samp>, optionally including 32-bit files
+(<samp>R-3.3.1patched-win32.msi</samp>) and from a 64-bit build of R
+(<samp>R-3.3.1patched-win64.msi</samp>, optionally including 32-bit files
 by setting the macro <code>HOME32</code>, when the name is
-<samp>R-3.3.1-win.msi</samp>).  Unlike the main installer, a 64-bit
+<samp>R-3.3.1patched-win.msi</samp>).  Unlike the main installer, a 64-bit
 MSI installer can only be run on 64-bit Windows.
 </p>
 <p>Thanks to David del Campo (Dept of Statistics, University of Oxford)
@@ -1859,6 +1912,9 @@
 <a name="index-OS-X-1"></a>
 
 
+<p>(&lsquo;OS X&rsquo; will be renamed to &lsquo;macOS&rsquo; as from version 10.12 &lsquo;macOS Sierra&rsquo;
+in late 2016.)
+</p>
 <p>The front page of a <acronym>CRAN</acronym> site has a link &lsquo;Download R for OS
 X&rsquo;. Click on that, then download the file <samp>R-3.3.1.pkg</samp>
 and install it. This runs on OS X 10.9 and later (Mavericks, Yosemite,
@@ -1921,7 +1977,7 @@
 </p>
 
 <p>You can run command-line R and <code>Rscript</code> from a
-Terminal<a name="DOCF13" href="#FOOT13"><sup>13</sup></a> so these can be
+Terminal<a name="DOCF16" href="#FOOT16"><sup>16</sup></a> so these can be
 typed as commands like any other Unix-alike: see the next chapter of
 this manual.  There are some small differences which may surprise users
 of R on other platforms, notably the default location of the personal
@@ -2133,7 +2189,7 @@
 </p>
 <p>R comes with a single library, <samp><var>R_HOME</var>/library</samp> which is
 the value of the R object &lsquo;<samp>.Library</samp>&rsquo; containing the standard and
-recommended<a name="DOCF14" href="#FOOT14"><sup>14</sup></a>  packages.
+recommended<a name="DOCF17" href="#FOOT17"><sup>17</sup></a>  packages.
 Both sites and users can create others and make use of them (or not) in
 an R session.  At the lowest level &lsquo;<samp>.libPaths()</samp>&rsquo; can be used to
 add paths to the collection of libraries or to report the current
@@ -2144,7 +2200,7 @@
 <p>R will automatically make use of a site-specific library
 <samp><var>R_HOME</var>/site-library</samp> if this exists (it does not in a
 vanilla R installation).  This location can be overridden by
-setting<a name="DOCF15" href="#FOOT15"><sup>15</sup></a> &lsquo;<samp>.Library.site</samp>&rsquo; in
+setting<a name="DOCF18" href="#FOOT18"><sup>18</sup></a> &lsquo;<samp>.Library.site</samp>&rsquo; in
 <samp><var>R_HOME</var>/etc/Rprofile.site</samp>, or (not recommended) by setting
 the
 <a name="index-R_005fLIBS_005fSITE"></a>
@@ -2191,7 +2247,7 @@
 <p>For most users it suffices to call
 &lsquo;<samp>install.packages(<var>pkgname</var>)</samp>&rsquo; or its GUI equivalent if the
 intention is to install a <acronym>CRAN</acronym> package and internet access is
-available.<a name="DOCF16" href="#FOOT16"><sup>16</sup></a>  On most systems &lsquo;<samp>install.packages()</samp>&rsquo;
+available.<a name="DOCF19" href="#FOOT19"><sup>19</sup></a>  On most systems &lsquo;<samp>install.packages()</samp>&rsquo;
 will allow packages to be selected from a list box (typically with
 several thousand items).
 </p>
@@ -2309,7 +2365,7 @@
 architecture was successfully installed.  The exceptions are packages
 with a non-empty <samp>configure.win</samp> script or which make use of
 <samp>src/Makefile.win</samp>.  If <samp>configure.win</samp> does something
-appropriate to both architectures use<a name="DOCF17" href="#FOOT17"><sup>17</sup></a> option
+appropriate to both architectures use<a name="DOCF20" href="#FOOT20"><sup>20</sup></a> option
 <samp>--force-biarch</samp>: otherwise <code>R CMD INSTALL
 --merge-multiarch</code> can be applied to a source tarball to merge separate
 32- and 64-bit installs.  (This can only be applied to a tarball, and
@@ -2333,7 +2389,7 @@
 on a Unix-alike.  All of these can be set in
 <samp>src/gnuwin32/MkRules.local</samp> when R is built from sources (see
 the comments in <samp>src/gnuwin32/MkRules.dist</samp>), or in
-file<a name="DOCF18" href="#FOOT18"><sup>18</sup></a>
+file<a name="DOCF21" href="#FOOT21"><sup>21</sup></a>
 <samp>etc/i386/Makeconf</samp> or <samp>etc/x64/Makeconf</samp> for an installed
 version of R.  In the latter case only <code>BINPREF</code> is used, with
 the 64 bit path used in <samp>etc/x64/Makeconf</samp>. The version used by
@@ -2555,7 +2611,7 @@
 </p>
 <p>On OS X the primary mechanism is to embed the absolute path to dependent
 dynamic libraries into an object when it is compiled.  Few R packages
-arrange to do so, but it can be edited<a name="DOCF19" href="#FOOT19"><sup>19</sup></a> <em>via</em> <code>install_name_tool</code> &mdash;
+arrange to do so, but it can be edited<a name="DOCF22" href="#FOOT22"><sup>22</sup></a> <em>via</em> <code>install_name_tool</code> &mdash;
 that only deals with direct dependencies and those would also need to be
 compiled to include the absolute paths of their dependencies.  If the
 choice of absolute path is to be deferred to load time, how they are
@@ -2837,7 +2893,7 @@
 <a name="Locales-under-Unix_002dalikes-1"></a>
 <h4 class="subsection">7.1.1 Locales under Unix-alikes</h4>
 
-<p>Modern Linux uses the XPG<a name="DOCF20" href="#FOOT20"><sup>20</sup></a> locale specifications which have the form
+<p>Modern Linux uses the XPG<a name="DOCF23" href="#FOOT23"><sup>23</sup></a> locale specifications which have the form
 &lsquo;<samp>en_GB</samp>&rsquo;, &lsquo;<samp>en_GB.UTF-8</samp>&rsquo;, &lsquo;<samp>aa_ER.UTF-8@saaho</samp>&rsquo;,
 &lsquo;<samp>de_AT.iso885915@euro</samp>&rsquo;, the components being in the order listed
 above.  (See <code>man locale</code> and <code>locale -a</code> for more
@@ -2898,7 +2954,7 @@
 <a name="index-LC_005fALL"></a>
 <a name="index-LC_005fMESSAGES"></a>
 <a name="index-LANG"></a>
-<code>LANGUAGE</code> and then<a name="DOCF21" href="#FOOT21"><sup>21</sup></a>
+<code>LANGUAGE</code> and then<a name="DOCF24" href="#FOOT24"><sup>24</sup></a>
 by the environment variables <code>LC_ALL</code>, <code>LC_MESSAGES</code> and
 <code>LANG</code>. (The last three are normally used to set the locale and so
 should not be needed, but the first is only used to select the language
@@ -2912,7 +2968,7 @@
 <p>It is usually possible to change the language once R is running
 <em>via</em> (not Windows) <code>Sys.setlocale(&quot;LC_MESSAGES&quot;,
 &quot;new_locale&quot;)</code>, or by setting an environment variable such as
-<code>LANGUAGE</code>, <em>provided</em><a name="DOCF22" href="#FOOT22"><sup>22</sup></a> the language you are changing to can be output in
+<code>LANGUAGE</code>, <em>provided</em><a name="DOCF25" href="#FOOT25"><sup>25</sup></a> the language you are changing to can be output in
 the current character set.  But this is OS-specific, and has been known
 to stop working on an OS upgrade.
 </p>
@@ -2951,7 +3007,7 @@
 specific translation catalogue exists but does not contain a
 translation, the less specific catalogues are consulted.  For example,
 R has catalogues for &lsquo;<samp>en_GB</samp>&rsquo; that translate the Americanisms
-(e.g., &lsquo;<samp>gray</samp>&rsquo;) in the standard messages into English.<a name="DOCF23" href="#FOOT23"><sup>23</sup></a>  Two other examples: there are catalogues
+(e.g., &lsquo;<samp>gray</samp>&rsquo;) in the standard messages into English.<a name="DOCF26" href="#FOOT26"><sup>26</sup></a>  Two other examples: there are catalogues
 for &lsquo;<samp>es</samp>&rsquo;, which is Spanish as written in Spain and these will by
 default also be used in Spanish-speaking Latin American countries, and
 also for &lsquo;<samp>pt_BR</samp>&rsquo;, which are used for Brazilian locales but not for
@@ -2967,7 +3023,7 @@
 quotes.
 </p>
 <p>If no suitable translation catalogue is found or a particular message is
-not translated in any suitable catalogue, &lsquo;English&rsquo;<a name="DOCF24" href="#FOOT24"><sup>24</sup></a> is used.
+not translated in any suitable catalogue, &lsquo;English&rsquo;<a name="DOCF27" href="#FOOT27"><sup>27</sup></a> is used.
 </p>
 <p>See <a href="https://developer.r-project.org/Translations30.html">https://developer.r-project.org/Translations30.html</a> for how to
 prepare and install translation catalogues.
@@ -2985,8 +3041,8 @@
 64-bit is.
 </p>
 <p>All current versions of R use 32-bit integers and
-<acronym>ISO</acronym>/<acronym>IEC</acronym>&nbsp;60559<a name="DOCF25" href="#FOOT25"><sup>25</sup></a> double-precision reals, and so compute to
-the same precision<a name="DOCF26" href="#FOOT26"><sup>26</sup></a> and with the same limits on the sizes of
+<acronym>ISO</acronym>/<acronym>IEC</acronym>&nbsp;60559<a name="DOCF28" href="#FOOT28"><sup>28</sup></a> double-precision reals, and so compute to
+the same precision<a name="DOCF29" href="#FOOT29"><sup>29</sup></a> and with the same limits on the sizes of
 numerical quantities.  The principal difference is in the size of the
 pointers.
 </p>
@@ -3049,7 +3105,7 @@
 <h2 class="chapter">9 The standalone Rmath library</h2>
 
 <p>The routines supporting the distribution and
-special<a name="DOCF27" href="#FOOT27"><sup>27</sup></a> functions in R
+special<a name="DOCF30" href="#FOOT30"><sup>30</sup></a> functions in R
 and a few others are declared in C header file <samp>Rmath.h</samp>.  These
 can be compiled into a standalone library for linking to other
 applications.  (Note that they are not a separate library when R is
@@ -3177,7 +3233,7 @@
 <a name="Windows-1"></a>
 <h3 class="section">9.2 Windows</h3>
 
-<p>You need to set up<a name="DOCF28" href="#FOOT28"><sup>28</sup></a> almost all the
+<p>You need to set up<a name="DOCF31" href="#FOOT31"><sup>31</sup></a> almost all the
 tools to make R and then run (in a Unix-like shell)
 </p>
 <div class="example">
@@ -3268,11 +3324,11 @@
 <h3 class="section">A.1 Essential programs and libraries</h3>
 
 <p>You need a means of compiling C and FORTRAN 90 (see <a href="#Using-FORTRAN">Using FORTRAN</a>).  Your C compiler should be
-<acronym>ISO</acronym>/<acronym>IEC</acronym>&nbsp;60059<a name="DOCF29" href="#FOOT29"><sup>29</sup></a>, POSIX 1003.1 and C99-compliant.<a name="DOCF30" href="#FOOT30"><sup>30</sup></a>  R tries to choose suitable flags for
+<acronym>ISO</acronym>/<acronym>IEC</acronym>&nbsp;60059<a name="DOCF32" href="#FOOT32"><sup>32</sup></a>, POSIX 1003.1 and C99-compliant.<a name="DOCF33" href="#FOOT33"><sup>33</sup></a>  R tries to choose suitable flags for
 the C compilers it knows about, but you may have to set <code>CC</code> or
 <code>CFLAGS</code> suitably.  For many versions of <code>gcc</code> with
 <code>glibc</code> this means including
-<samp>-std=gnu99</samp><a name="DOCF31" href="#FOOT31"><sup>31</sup></a>.  (Note that options essential to
+<samp>-std=gnu99</samp><a name="DOCF34" href="#FOOT34"><sup>34</sup></a>.  (Note that options essential to
 run the compiler even for linking, such as those to set the
 architecture, should be specified as part of <code>CC</code> rather than in
 <code>CFLAGS</code>.)
@@ -3296,21 +3352,21 @@
 <code>&quot;latin1&quot;</code> and <code>&quot;UTF-8&quot;</code>, to recognize <code>&quot;&quot;</code> (as the
 current encoding) and <code>&quot;ASCII&quot;</code>, and to translate to and from the
 Unicode wide-character formats <code>&quot;UCS-[24][BL]E&quot;</code> &mdash; this is true
-by default for <code>glibc</code><a name="DOCF32" href="#FOOT32"><sup>32</sup></a> but not of most commercial Unixes.  However, you
+by default for <code>glibc</code><a name="DOCF35" href="#FOOT35"><sup>35</sup></a> but not of most commercial Unixes.  However, you
 can make use of <acronym>GNU</acronym> <code>libiconv</code> (as used on OS X: see
 <a href="https://www.gnu.org/software/libiconv/">https://www.gnu.org/software/libiconv/</a>).
 </p>
-<p>The OS needs to have enough support<a name="DOCF33" href="#FOOT33"><sup>33</sup></a> for wide-character
+<p>The OS needs to have enough support<a name="DOCF36" href="#FOOT36"><sup>36</sup></a> for wide-character
 types: this is checked at configuration.  A small number of POSIX
-functions<a name="DOCF34" href="#FOOT34"><sup>34</sup></a> are essential, and others<a name="DOCF35" href="#FOOT35"><sup>35</sup></a> will be used if available.
+functions<a name="DOCF37" href="#FOOT37"><sup>37</sup></a> are essential, and others<a name="DOCF38" href="#FOOT38"><sup>38</sup></a> will be used if available.
 </p>
 <p>Installations of <code>zlib</code> (version 1.2.5 or later), <code>libbz2</code>
 (version 1.0.6 or later: called <strong>bzip2-libs</strong>/<strong>bzip2-devel</strong> or
 <strong>libbz2-1.0</strong>/<strong>libbz2-dev</strong> by some Linux distributions),
-<code>liblzma</code><a name="DOCF36" href="#FOOT36"><sup>36</sup></a> version 5.0.3 or
+<code>liblzma</code><a name="DOCF39" href="#FOOT39"><sup>39</sup></a> version 5.0.3 or
 later are required.
 </p>
-<p>PCRE<a name="DOCF37" href="#FOOT37"><sup>37</sup></a> (version 8.32 or later, although versions 8.10&ndash;8.31 will
+<p>PCRE<a name="DOCF40" href="#FOOT40"><sup>40</sup></a> (version 8.32 or later, although versions 8.10&ndash;8.31 will
 be accepted with a deprecation warning) is required (or just its library
 and headers if packaged separately).  PCRE must be built with UTF-8
 support (not the default, and checked by <code>configure</code>) and
@@ -3324,13 +3380,13 @@
 </pre></div>
 <p>The <samp>--enable-jit</samp> flag is supported for most common CPUs.
 </p>
-<p>Library <code>libcurl</code> (version 7.28.0 or later) is required.
-Information on <code>libcurl</code> is found from the <code>curl-config</code>
-script: if that is missing or needs to be overridden<a name="DOCF38" href="#FOOT38"><sup>38</sup></a> there are macros to do so described in file
-<samp>config.site</samp>.
+<p>Library <code>libcurl</code> (version 7.28.0 or later<a name="DOCF41" href="#FOOT41"><sup>41</sup></a>) is required.  Information on <code>libcurl</code> is
+found from the <code>curl-config</code> script: if that is missing or needs
+to be overridden<a name="DOCF42" href="#FOOT42"><sup>42</sup></a> there are macros to
+do so described in file <samp>config.site</samp>.
 </p>
 <p>A <code>tar</code> program is needed to unpack the sources and packages
-(including the recommended packages).  A version<a name="DOCF39" href="#FOOT39"><sup>39</sup></a> that can
+(including the recommended packages).  A version<a name="DOCF43" href="#FOOT43"><sup>43</sup></a> that can
 automagically detect compressed archives is preferred for use with
 <code>untar()</code>: the configure script looks for <code>gtar</code> and
 <code>gnutar</code> before
@@ -3350,7 +3406,7 @@
 <samp>texinfo.tex</samp> installed (which is part of the <acronym>GNU</acronym>
 <strong>texinfo</strong> distribution but is often made part of the TeX package
 in re-distributions) as well as
-<code>texi2dvi</code>.<a name="DOCF40" href="#FOOT40"><sup>40</sup></a>
+<code>texi2dvi</code>.<a name="DOCF44" href="#FOOT44"><sup>44</sup></a>
 Further, the versions of <code>texi2dvi</code> and <samp>texinfo.tex</samp> need
 to be compatible: we have seen problems with older TeX distributions.
 </p>
@@ -3384,6 +3440,15 @@
 <p>The essential programs should be in your <code>PATH</code> at the time
 <code>configure</code> is run: this will capture the full paths.
 </p>
+<p>Those distributing binary versions of R may need to be aware of the
+licences of the external libraries it is linked to (including &lsquo;useful&rsquo;
+libraries from the next section).  The <code>liblzma</code> library is in the
+public domain and X11, <code>libbzip2</code>, <code>libcurl</code> and <code>zlib</code>
+have MIT-style licences. PCRE has a BSD-style licence which requires
+distribution of the licence (included in R&rsquo;s <samp>COPYRIGHTS</samp> file)
+in binary distributions.  GNU <code>readline</code> is licensed under GPL
+(which version(s) depending on the <code>readline</code> version).
+</p>
 <hr>
 <a name="Useful-libraries-and-programs"></a>
 <a name="Useful-libraries-and-programs-1"></a>
@@ -3404,18 +3469,18 @@
 &lsquo;<samp>pangocairo</samp>&rsquo; package is installed (and if not, &lsquo;<samp>cairo</samp>&rsquo;) and if
 additional flags are needed for the &lsquo;<samp>cairo-xlib</samp>&rsquo; package, then if
 suitable code can be compiled.  These tests will fail if
-<code>pkg-config</code> is not installed<a name="DOCF41" href="#FOOT41"><sup>41</sup></a>, and are likely to fail if <code>cairo</code> was built
+<code>pkg-config</code> is not installed<a name="DOCF45" href="#FOOT45"><sup>45</sup></a>, and are likely to fail if <code>cairo</code> was built
 statically (unusual).  Most systems with <code>Gtk+</code> 2.8 or later
 installed will have suitable libraries
 </p>
 <p>For the best font experience with these devices you need suitable fonts
 installed: Linux users will want the <code>urw-fonts</code> package.  On
 platforms which have it available, the <code>msttcorefonts</code>
-package<a name="DOCF42" href="#FOOT42"><sup>42</sup></a> provides
+package<a name="DOCF46" href="#FOOT46"><sup>46</sup></a> provides
 TrueType versions of Monotype fonts such as Arial and Times New Roman.
 Another useful set of fonts is the &lsquo;liberation&rsquo; TrueType fonts available
 at
-<a href="https://fedorahosted.org/liberation-fonts/">https://fedorahosted.org/liberation-fonts/</a>,<a name="DOCF43" href="#FOOT43"><sup>43</sup></a> which cover the Latin, Greek and Cyrillic alphabets
+<a href="https://fedorahosted.org/liberation-fonts/">https://fedorahosted.org/liberation-fonts/</a>,<a name="DOCF47" href="#FOOT47"><sup>47</sup></a> which cover the Latin, Greek and Cyrillic alphabets
 plus a fair range of signs.  These share metrics with Arial, Times New
 Roman and Courier New, and contain fonts rather similar to the first two
 (<a href="https://en.wikipedia.org/wiki/Liberation_fonts">https://en.wikipedia.org/wiki/Liberation_fonts</a>).  Then there
@@ -3447,13 +3512,6 @@
 at <a href="https://github.com/laurikari/tre/">https://github.com/laurikari/tre/</a>, but at the time of writing
 the resulting build will not pass its checks.).
 </p>
-<p>Library <code>liblzma</code> from <code>xz-utils</code> version 5.0.3 or later
-(including 5.2.x) will be used if installed: the version in the R
-sources can be selected instead by configuring with
-<samp>--without-system-xz</samp>.  Systems differ in what they call the
-package including this: e.g. on Fedora the library is in
-&lsquo;<samp>xz-libs</samp>&rsquo; and the headers in &lsquo;<samp>xz-devel</samp>&rsquo;.
-</p>
 <p>An implementation of <acronym>XDR</acronym> is required, and the R sources
 contain one which is likely to suffice (although a system version may
 have higher performance).  <acronym>XDR</acronym> is part of <acronym>RPC</acronym> and
@@ -3520,7 +3578,7 @@
 <p>Versions of Tcl/Tk up to 8.5.19 and 8.6.4 have been tested (including
 most versions of 8.4.x, but not recently).
 </p>
-<p>Note that the <samp>tk.h</samp> header includes<a name="DOCF44" href="#FOOT44"><sup>44</sup></a> X11 headers, so you will need X11 and its
+<p>Note that the <samp>tk.h</samp> header includes<a name="DOCF48" href="#FOOT48"><sup>48</sup></a> X11 headers, so you will need X11 and its
 development files installed.
 </p>
 <hr>
@@ -3588,10 +3646,10 @@
 most cases this will be a C++98 compiler, and as from R 3.1.0 it is
 possible to specify an alternative compiler for use with C++11 by the
 configure variables <code>CXX1X</code>, <code>CXX1XSTD</code>, <code>CXX1XFLAGS</code> and
-similar.  Again, <code>configure</code> will normally find a suitable value
-for <code>CXX1XSTD</code> if the compiler given by <code>CXX</code> is capable of
-compiling C++11 code, but it is possible that a completely different
-compiler will be needed.
+similar (see <a href="#C_002b_002b-Support">C++ Support</a>).  Again, <code>configure</code> will normally
+find a suitable value for <code>CXX1XSTD</code> if the compiler given by
+<code>CXX</code> is capable of compiling C++11 code, but it is possible that a
+completely different compiler will be needed.
 </p>
 <p>Other packages need full Fortran 90 (or later) support.  For source
 files with extension <samp>.f90</samp> or <samp>.f95</samp>, the compiler defined by
@@ -3651,7 +3709,7 @@
 position-independent code), and that is not checked.
 </p>
 <p>Some enhanced <acronym>BLAS</acronym>es are compiler-system-specific
-(<code>sunperf</code> on Solaris<a name="DOCF45" href="#FOOT45"><sup>45</sup></a>, <code>libessl</code> on IBM,
+(<code>sunperf</code> on Solaris<a name="DOCF49" href="#FOOT49"><sup>49</sup></a>, <code>libessl</code> on IBM,
 <code>Accelerate</code> on OS X).  The correct incantation for these is often
 found <em>via</em> <samp>--with-blas</samp> with no value on the appropriate
 platforms.
@@ -3733,7 +3791,7 @@
 <a name="ACML-1"></a>
 <h4 class="subsubsection">A.3.1.2 ACML</h4>
 
-<p>For &lsquo;<samp>x86_64</samp>&rsquo; processors<a name="DOCF46" href="#FOOT46"><sup>46</sup></a>  under Linux there is the AMD Core Math Library (ACML).
+<p>For &lsquo;<samp>x86_64</samp>&rsquo; processors<a name="DOCF50" href="#FOOT50"><sup>50</sup></a>  under Linux there is the AMD Core Math Library (ACML).
 For the <code>gcc</code> version we could use
 </p>
 <div class="example">
@@ -4056,7 +4114,7 @@
 the main R executable <samp>R.bin</samp> against that library.  This can
 only be done if all the code (including system libraries) can be
 compiled into a dynamic library, and there may be a
-performance<a name="DOCF47" href="#FOOT47"><sup>47</sup></a> penalty.  So you probably
+performance<a name="DOCF51" href="#FOOT51"><sup>51</sup></a> penalty.  So you probably
 only want this if you will be using an application which embeds R.
 Note that C code in packages installed on an R system linked with
 <samp>--enable-R-shlib</samp> is linked against the dynamic library and so
@@ -4073,7 +4131,7 @@
 by option <samp>--with-system-valgrind-headers</samp>: they will be used if
 present (on Linux they may be in a separate package such as
 <strong>valgrind-devel</strong>).  Note though that there is no guarantee that the
-code in R will be compatible with very old<a name="DOCF48" href="#FOOT48"><sup>48</sup></a> or future <code>valgrind</code>
+code in R will be compatible with very old<a name="DOCF52" href="#FOOT52"><sup>52</sup></a> or future <code>valgrind</code>
 headers.
 </p>
 <p>If you need to re-configure R with different options you may need to run
@@ -4277,7 +4335,7 @@
 <code>epcf90</code>,
 <code>g77</code>, <code>f77</code>, <code>xlf</code>, <code>frt</code>,
 <code>pgf77</code>, <code>cf77</code>, <code>fort77</code>, <code>fl32</code>,
-<code>af77</code> (in that order)<a name="DOCF49" href="#FOOT49"><sup>49</sup></a>, and use whichever is found first; if none is found,
+<code>af77</code> (in that order)<a name="DOCF53" href="#FOOT53"><sup>53</sup></a>, and use whichever is found first; if none is found,
 R cannot be compiled.
 However, if <code>CC</code> is <code>gcc</code>, the matching FORTRAN compiler
 (<code>g77</code> for <code>gcc</code>&nbsp;3 and <code>gfortran</code> for
@@ -4306,7 +4364,7 @@
 </p>
 <p>Some of the FORTRAN code makes use of <code>COMPLEX*16</code> variables, which
 is a Fortran 90 extension.  This is checked for at configure
-time<a name="DOCF50" href="#FOOT50"><sup>50</sup></a>, but you may need to avoid
+time<a name="DOCF54" href="#FOOT54"><sup>54</sup></a>, but you may need to avoid
 compiler flags asserting FORTRAN 77 compliance.
 </p>
 <p>Compiling the version of LAPACK in the R sources also requires some
@@ -4525,7 +4583,7 @@
 </p>
 <p>When X11 was designed, most displays were around 75dpi, whereas today
 they are of the order of 100dpi or more.  If you find that X11()
-is reporting<a name="DOCF51" href="#FOOT51"><sup>51</sup></a> missing font sizes, especially larger ones, it is likely
+is reporting<a name="DOCF55" href="#FOOT55"><sup>55</sup></a> missing font sizes, especially larger ones, it is likely
 that you are not using scalable fonts and have not installed the 100dpi
 versions of the X11 fonts.  The names and details differ by system, but
 will likely have something like Fedora&rsquo;s
@@ -4643,7 +4701,7 @@
 the best possible performance on the machine on which R is being
 installed: if the compilation is for a site-wide installation, it may
 still be desirable to use something like
-<samp>-mtume=core2</samp>.<a name="DOCF52" href="#FOOT52"><sup>52</sup></a> It is also possible to increase the
+<samp>-mtume=core2</samp>.<a name="DOCF56" href="#FOOT56"><sup>56</sup></a> It is also possible to increase the
 optimization levels to <samp>-O3</samp>: however for many versions of the
 compilers this has caused problems in at least one <acronym>CRAN</acronym>
 package.
@@ -4719,7 +4777,7 @@
 R packages only the variant using <code>libcxxabi</code> was successful.
 </p>
 <p>Most builds of <code>clang</code> have no OpenMP support.  Builds of
-version 3.7.0 or later may.<a name="DOCF53" href="#FOOT53"><sup>53</sup></a>
+version 3.7.0 or later may.<a name="DOCF57" href="#FOOT57"><sup>57</sup></a>
 </p>
 <hr>
 <a name="Intel-compilers"></a>
@@ -4774,52 +4832,17 @@
 FCFLAGS=$F77FLAGS
 </pre></div>
 
-
-<hr>
-<a name="Oracle-Solaris-Studio-compilers"></a>
-<a name="Oracle-Solaris-Studio-compilers-1"></a>
-<h4 class="subsection">C.2.3 Oracle Solaris Studio compilers</h4>
-
-<p>Brian Ripley tested the Sun Studio 12 compilers, since renamed to Oracle
-Solaris Studio.  On &lsquo;<samp>x86_64</samp>&rsquo; Linux with
-</p>
-<div class="example">
-<pre class="example">CC=suncc
-CFLAGS=&quot;-xO5 -xc99 -xlibmil -nofstore&quot;
-CPICFLAGS=-Kpic
-F77=sunf95
-FFLAGS=&quot;-O5 -libmil -nofstore&quot;
-FPICFLAGS=-Kpic
-CXX=&quot;sunCC -library=stlport4&quot;
-CXXFLAGS=&quot;-xO5 -xlibmil -nofstore -features=tmplrefstatic&quot;
-CXXPICFLAGS=-Kpic
-FC=sunf95
-FCFLAGS=$FFLAGS
-FCPICFLAGS=-Kpic
-LDFLAGS=-L/opt/sunstudio12.1/rtlibs/amd64
-SHLIB_LDFLAGS=-shared
-SHLIB_CXXLDFLAGS=-G
-SHLIB_FCLDFLAGS=-G
-SAFE_FFLAGS=&quot;-O5 -libmil&quot;
-</pre></div>
-
-<p><samp>-m64</samp> could be added, but was the default.  Do not use
-<samp>-fast</samp>: see the warnings under Solaris. (The C++ options are
-also explained under Solaris.)
-</p>
-<p>Others have found on at least some versions of &lsquo;<samp>ix86</samp>&rsquo; Linux that
-the configure flag <samp>--disable-largefile</samp> was needed (since
-<samp>glob.h</samp> on that platform presumed <code>gcc</code> was being used).
-</p>
-
 <hr>
 <a name="OS-X"></a>
 <a name="OS-X-2"></a>
 <h3 class="section">C.3 OS X</h3>
 <a name="index-OS-X-2"></a>
 
+<p>(&lsquo;OS X&rsquo; will be renamed to &lsquo;macOS&rsquo; from version 10.12 &lsquo;macOS Sierra&rsquo; in
+late 2016.)
+</p>
 <p>The instructions here are for &lsquo;<samp>x86_64</samp>&rsquo; builds on 10.9
-(Mavericks) or later.  In principle<a name="DOCF54" href="#FOOT54"><sup>54</sup></a> R can be
+(Mavericks) or later.  In principle<a name="DOCF58" href="#FOOT58"><sup>58</sup></a> R can be
 built for 10.6 to 10.8 but these has not been tested recently.
 </p>
 <p>To build R you need Apple&rsquo;s &lsquo;Command Line Tools&rsquo;: these can be
@@ -4829,10 +4852,11 @@
 Xcode, this provides the command-line tools.  The tools will need to be
 reinstalled when OS X is upgraded, as upgrading partially removes them.)
 </p>
-<p>You need <code>readline</code><a name="DOCF55" href="#FOOT55"><sup>55</sup></a> and a Fortran compiler.  Those and other binary
-components are available from <a href="https://r.research.att.com/libs">https://r.research.att.com/libs</a>:
-you are likely to need <code>pcre</code> and <code>xz</code> (recent OS X provides
-libraries but not headers for these).
+<p>You need <code>readline</code><a name="DOCF59" href="#FOOT59"><sup>59</sup></a> and a Fortran compiler.
+Those and other binary components are available from
+<a href="https://r.research.att.com/libs">https://r.research.att.com/libs</a>: you are likely to need
+<code>pcre</code> and <code>xz</code> (recent OS X provides libraries but not
+headers for these).
 </p>
 <p>An X sub-system is required unless configuring using
 <samp>--without-x</samp>: see <a href="https://xquartz.macosforge.org/">https://xquartz.macosforge.org/</a>.
@@ -4841,7 +4865,7 @@
 <samp>--with-aqua</samp> (which is the default): <code>quartz()</code> then
 becomes the default device when running R at the console and X11
 would only be used for the command-line-R data editor/viewer and one
-version of Tcl/Tk.  (This needs an Objective-C compiler<a name="DOCF56" href="#FOOT56"><sup>56</sup></a> which can compile the source code of
+version of Tcl/Tk.  (This needs an Objective-C compiler<a name="DOCF60" href="#FOOT60"><sup>60</sup></a> which can compile the source code of
 <code>quartz()</code>.)
 </p>
 <p>Use <samp>--without-aqua</samp> if you want a standard Unix-alike build:
@@ -4867,7 +4891,7 @@
 
 <p>with <code>clang</code> and <code>clang++</code> from the &lsquo;Command Line
 Tools&rsquo; and the Fortran compiler from
-<a href="https://r.research.att.com/libs/gfortran-4.8.2-darwin13.tar.bz2">https://r.research.att.com/libs/gfortran-4.8.2-darwin13.tar.bz2</a>.<a name="DOCF57" href="#FOOT57"><sup>57</sup></a>
+<a href="https://r.research.att.com/libs/gfortran-4.8.2-darwin13.tar.bz2">https://r.research.att.com/libs/gfortran-4.8.2-darwin13.tar.bz2</a>.<a name="DOCF61" href="#FOOT61"><sup>61</sup></a>
 </p>
 <p>Other builds of <code>gfortran</code> are available: see
 <a href="https://gcc.gnu.org/wiki/GFortranBinaries">https://gcc.gnu.org/wiki/GFortranBinaries</a> and
@@ -4878,7 +4902,8 @@
 <p>More recent and complete distributions of <code>clang</code> are usually
 available from <a href="http://llvm.org/releases/">http://llvm.org/releases/</a>.  In particular, these
 include support for the &lsquo;Address Sanitizer&rsquo; (not included by Apple until
-Xcode 7) and for OpenMP<a name="DOCF58" href="#FOOT58"><sup>58</sup></a> in version 3.7.0 and later.
+Xcode 7) and for OpenMP<a name="DOCF62" href="#FOOT62"><sup>62</sup></a> in version
+3.7.0 and later.
 </p>
 <p>Pre-compiled versions of many of the <a href="#Useful-libraries-and-programs">Useful libraries and programs</a>
 are available from <a href="https://r.research.att.com/libs/">https://r.research.att.com/libs/</a>.  You will
@@ -4910,7 +4935,7 @@
 </pre></div>
 
 <p>to provide potentially higher-performance versions of the <acronym>BLAS</acronym>
-and LAPACK routines.<a name="DOCF59" href="#FOOT59"><sup>59</sup></a>
+and LAPACK routines.<a name="DOCF63" href="#FOOT63"><sup>63</sup></a>
 </p>
 <p>Looking at the top of
 <samp>/Library/Frameworks/R.framework/Resources/etc/Makeconf</samp>
@@ -4928,9 +4953,10 @@
 </p>
 <p>The TeX implementation used by the developers is MacTeX
 (<a href="https://www.tug.org/mactex/">https://www.tug.org/mactex/</a>): the full installation is about
-4GB, but a smaller version is available at
+5GB, but a smaller version is available at
 <a href="https://www.tug.org/mactex/morepackages.html">https://www.tug.org/mactex/morepackages.html</a>: you will need to
-add some packages, e.g. for the 2015 version we needed to add
+add some packages, e.g. for the 2016 version we needed to
+add<a name="DOCF64" href="#FOOT64"><sup>64</sup></a>
 <strong>cm-super</strong>, <strong>helvetic</strong>, <strong>inconsolata</strong> and <strong>texinfo</strong>
 which brought this to about 410MB.  &lsquo;<samp>TeX Live Utility</samp>&rsquo; (available
 <em>via</em> the MacTeX front page) provides a graphical means to manage
@@ -4953,14 +4979,15 @@
 <a name="El-Capitan-1"></a>
 <h4 class="subsection">C.3.1 El Capitan</h4>
 
-<p>There are problems resulting from the new-to-El-Capitan restriction that
+<p>There have been problems resulting from the new-to-El-Capitan restriction that
 only Apple is allowed to install software under <samp>/usr</samp>: this
-affects <em>inter alia</em> <code>MacTeX</code> and <code>XQuartz</code>.  For
-<code>MacTeX</code> it is necessary to include <samp>/Library/TeX/texbin</samp> in
-your path rather than <samp>/usr/texbin</samp>.  Upgrading will move
-disallowed files to under <samp>/Library/SystemMigration/usr</samp>: this
-includes <samp>/usr/X11R6</samp>, <samp>/usr/texbin</samp>, <samp>/usr/bin/R</samp>,
-<samp>/usr/bin/Rscript</samp> but not the link <samp>/usr/X11</samp>.
+affects <em>inter alia</em> <code>MacTeX</code> and <code>XQuartz</code> which changed
+their installation locations.  For <code>MacTeX</code> it is necessary to
+include <samp>/Library/TeX/texbin</samp> in your path rather than
+<samp>/usr/texbin</samp>.  Upgrading will move disallowed files to under
+<samp>/Library/SystemMigration/usr</samp>: this includes <samp>/usr/X11R6</samp>,
+<samp>/usr/texbin</samp>, <samp>/usr/bin/R</samp>, <samp>/usr/bin/Rscript</samp> but not
+the link <samp>/usr/X11</samp>.
 </p>
 <p><code>configure</code> can be told to look for <code>X11</code> in
 <code>XQuartz</code>&rsquo;s main location of <samp>/opt/X11</samp>, e.g. by
@@ -4968,8 +4995,7 @@
 <pre class="example">--x-includes=/opt/X11/include --x-libraries=/opt/X11/lib
 </pre></div>
 
-<p>although the linked versions under <samp>/usr/X11</samp> will be found (if the
-link is present).
+<p>although linked versions under <samp>/usr/X11</samp> will be found.
 </p>
 <hr>
 <a name="Tcl_002fTk-headers-and-libraries"></a>
@@ -5021,7 +5047,7 @@
 <a name="Java"></a>
 <h4 class="subsection">C.3.3 Java</h4>
 
-<p>The situation with Java support on OS X is messy.<a name="DOCF60" href="#FOOT60"><sup>60</sup></a>
+<p>The situation with Java support on OS X is messy.<a name="DOCF65" href="#FOOT65"><sup>65</sup></a>
 </p>
 <p>OS X no longer comes with an installed Java runtime (JRE), and an OS X
 upgrade removes one if already installed: it is intended to be installed
@@ -5031,7 +5057,7 @@
 latest Java from Oracle (currently from
 <a href="http://www.oracle.com/technetwork/java/javase/downloads/index.html">http://www.oracle.com/technetwork/java/javase/downloads/index.html</a>).
 </p>
-<p>You may need to install what Apple calls &lsquo;legacy Java&rsquo;<a name="DOCF61" href="#FOOT61"><sup>61</sup></a> to suppress pop-up messages
+<p>You may need to install what Apple calls &lsquo;legacy Java&rsquo;<a name="DOCF66" href="#FOOT66"><sup>66</sup></a> to suppress pop-up messages
 even if you have a current version installed.
 </p>
 <p>To see what compatible versions of Java are currently installed, run
@@ -5136,15 +5162,17 @@
 <a name="index-Solaris"></a>
 
 <p>R has been built successfully on Solaris 10 (both Sparc and
-&lsquo;<samp>x86</samp>&rsquo;) using the (zero cost) Oracle Solaris Studio compilers:
-there has been some success with
-<code>gcc</code>&nbsp;4/<code>gfortran</code>.  (Recent Sun machines are AMD
-Opterons or Intel Xeons (&lsquo;<samp>amd64</samp>&rsquo;) rather than &lsquo;<samp>x86</samp>&rsquo;, but
-32-bit &lsquo;<samp>x86</samp>&rsquo; executables are the default.)
+&lsquo;<samp>x86</samp>&rsquo;) using the (zero cost) Oracle Developer
+Studio<a name="DOCF67" href="#FOOT67"><sup>67</sup></a> compilers: there has been some success with
+<code>gcc</code>/<code>gfortran</code>.  (Recent Sun machines are AMD Opterons
+or Intel Xeons (&lsquo;<samp>amd64</samp>&rsquo;) rather than &lsquo;<samp>x86</samp>&rsquo;, but 32-bit
+&lsquo;<samp>x86</samp>&rsquo; executables are the default.)  How these compilers
+identify<a name="DOCF68" href="#FOOT68"><sup>68</sup></a> themselves is slightly
+confusing: Solaris Studio versions 12.3 and 12.4 report versions 5.12
+and 5.13, and Developer Studio 12.5 reports 5.14.  We will only consider
+12.3 (Nov 2011) and later.
 </p>
 <p>There have been few reports on Solaris 11, with no known extra issues.
-Solaris 9 and earlier are now so old that it is unlikely that R is
-still used with them, and they will not be considered here.
 </p>
 <p>The Solaris versions of several of the tools needed to build R
 (e.g. <code>make</code>, <code>ar</code> and <code>ld</code>) are in
@@ -5187,8 +5215,9 @@
 you find time-zone abbreviations being given odd values (as has been
 seen on 64-bit builds without it).
 </p>
-<p>When using the Oracle compilers<a name="DOCF62" href="#FOOT62"><sup>62</sup></a> do <em>not</em> specify <samp>-fast</samp>, as this
-disables <acronym>IEEE</acronym> arithmetic and <code>make check</code> will fail.
+<p>When using the Oracle compilers do <em>not</em> specify <samp>-fast</samp>, as
+this disables <acronym>IEEE</acronym> arithmetic and <code>make check</code> will
+fail.
 </p>
 <p>It has been reported that some Solaris installations need
 </p>
@@ -5210,17 +5239,20 @@
 FFLAGS=-O
 CXX=&quot;CC -library=stlport4&quot;
 CXXFLAGS=-O
-FC=f95
+FC=$F77
 FCFLAGS=$FFLAGS
-FCLIBS=&quot;-lfai -lfsu&quot;
-R_LD_LIBRARY_PATH=&quot;/usr/local/lib:/opt/csw/gcc4/lib:/opt/csw/lib&quot;
+FCLIBS=&quot;-lfai -lfui -lfsu&quot;
+R_LD_LIBRARY_PATH=&quot;/usr/local/lib:/opt/csw/lib&quot;
 </pre></div>
 
+<p>(For version 12.5 of the compiler use <code>FCLIBS=-lfsu</code>.)
+</p>
 <p>For a 64-bit target add <samp>-m64</samp> to the compiler macros
-and use something like <code>LDFLAGS=-L/usr/local/lib/sparcv9</code> or
-<code>LDFLAGS=-L/usr/local/lib/amd64</code> as appropriate.
+and use something like <code>LDFLAGS=-L/usr/local/lib/amd64</code> or
+<code>LDFLAGS=-L/usr/local/lib/sparcv9</code> as appropriate (and other 64-bit
+library directories if used, e.g. <code>-L/opt/csw/lib/arm64</code>).
 It will also be necessary to point <code>pkg-config</code> at the 64-bit
-directories, e.g. one of
+directories, e.g. something like one of
 </p>
 <div class="example">
 <pre class="example">PKG_CONFIG_PATH=/opt/csw/lib/amd64/pkgconfig:/usr/lib/amd64/pkgconfig
@@ -5235,24 +5267,23 @@
 JAVA_LIBS=&quot;-L${JAVA_HOME}/lib/amd64/server \
   -R${JAVA_HOME}/lib/amd64/server -ljvm&quot;
 </pre></div>
-<p>With Solaris Studio 12.[23] on Sparc, <code>FCLIBS</code> needs to be
+<p>With version 12.3 on Sparc, <code>FCLIBS</code> needs to be
 </p>
 <div class="example">
-<pre class="example">FCLIBS=&quot;-lfai -lfai2 -lfsu&quot;
+<pre class="example">FCLIBS=&quot;-lfai -lfui -lfai2 -lfsu&quot;
 </pre></div>
 
 <p>(and possibly other Fortran libraries, but this suffices for the
 packages currently on CRAN).
 </p>
-<p>Currently &lsquo;<samp>amd64</samp>&rsquo; and &lsquo;<samp>sparcv9</samp>&rsquo; builds work
-out-of-the-box with Sun Studio 12u1 but not Solaris Studio 12.2 and
-12.3: <samp>libRblas.so</samp> and <samp>lapack.so</samp> are generated with code
-that causes relocation errors (which is being linked in from the Fortran
-libraries).  This means that building 64-bit R as a shared library
-may be impossible with Solaris Studio &gt;= 12.2.  For a standard build the
+<p>Builds of &lsquo;<samp>amd64</samp>&rsquo; and &lsquo;<samp>sparcv9</samp>&rsquo; stopped working at
+Solaris Studio 12.3: <samp>libRblas.so</samp> and <samp>lapack.so</samp> are
+generated with code that causes relocation errors (code which is being
+linked in from the Fortran libraries).  This means that building 64-bit
+R as a shared library may be impossible.  For a standard build the
 trick seems to be to manually set <code>FLIBS</code> to avoid the troublesome
 libraries.  For example, on &lsquo;<samp>amd64</samp>&rsquo; set in <samp>config.site</samp>
-something like
+something like (for 12.3)
 </p>
 <div class="example">
 <pre class="example">FLIBS_IN_SO=&quot;-R/opt/solarisstudio12.3/lib/amd64
@@ -5267,7 +5298,7 @@
  /opt/solarisstudio12.3/prod/lib/sparc/64/libfsu.so.1&quot;
 </pre></div>
 
-<p>By default the Solaris Studio compilers do not by default conform to the C99
+<p>By default the Oracle compilers do not by default conform to the C99
 standard (appendix F 8.9) on the return values of functions such as
 <code>log</code>: use <samp>-xlibmieee</samp> to ensure this.
 </p>
@@ -5275,7 +5306,7 @@
 performance: <samp>-xtarget=native</samp> (in <code>CFLAGS</code> etc) tunes the
 compilation to the current machine.
 </p>
-<p>Using <code>-xlibmil</code> in <code>CFLAGS</code> and <code>-xlibmil</code> in
+<p>Using <code>-xlibmil</code> in <code>CFLAGS</code> or <code>-xlibmil</code> in
 <code>FFLAGS</code> allows more system mathematical functions to be inlined.
 </p>
 
@@ -5292,7 +5323,7 @@
 packages (notably <a href="https://CRAN.R-project.org/package=mgcv"><strong>mgcv</strong></a> on &lsquo;<samp>x86</samp>&rsquo;) failed to compile at
 higher optimization levels with version 12.3.
 </p>
-<p>The Solaris Studio compilers provide several implementations of the
+<p>The Oracle compilers provide several implementations of the
 C++98 standard which select both the set of headers and a C++ runtime
 library.  These are selected by the <samp>-library</samp> flag, which as it
 is needed for both compiling and linking is best specified as part of
@@ -5305,14 +5336,6 @@
 needs to be.  Occasionally the option <samp>-library=stlport4,Crun</samp>
 has been needed.
 </p>
-<p>Several <acronym>CRAN</acronym> packages using C++ need the more liberal
-interpretation given by adding
-</p>
-<div class="example">
-<pre class="example">CXXFLAGS=&quot;-features=tmplrefstatic&quot;
-</pre></div>
-
-
 <a name="index-BLAS-library-2"></a>
 <a name="index-LAPACK-library-2"></a>
 
@@ -5348,64 +5371,77 @@
 </p>
 <p>The notes here are for <code>gcc</code> set up to use the Solaris linker:
 it can also be set up to use GNU <code>ld</code>, but that has not been
-tested.
+tested.  The tests were for compilers from the OpenCSW repository:
+Solaris systems often come with much older compilers installed under
+<samp>/usr/sfw/bin</samp> and in the past an explicit path such as
+<code>/opt/csw/gcc4/bin/gcc</code> was need to select the latest version of the
+compilers (which at the time of writing were linked from
+<samp>/opt/csw/bin</samp> to <samp>/opt/csw/gcc4/bin</samp>).  One of <samp>-m32</samp>
+or <samp>-m64</samp> will be the default and could be omitted, but it is not
+easy to find out which.
 </p>
-<p>Compilation for a 32-bit Sparc target with <code>gcc</code>&nbsp;4.9.2
-needed
+<p>Compilation for an &lsquo;<samp>x86</samp>&rsquo; target with <code>gcc</code>&nbsp;4.9.2 or
+<code>gcc</code>&nbsp;5.2.0 needed
 </p>
 <div class="example">
-<pre class="example">CPPFLAGS=-I/opt/csw/include
-LDFLAGS=&quot;-L/opt/csw/gcc4/lib -L/opt/csw/lib&quot;
+<pre class="example">CC=&quot;gcc -m32&quot;
+CPPFLAGS=&quot;-I/opt/csw/include -I/usr/local/include&quot;
+F77=&quot;gfortran -m32&quot;
+CXX=&quot;g++ -m32&quot;
+FC=&quot;gfortran -m32&quot;
+LDFLAGS=&quot;-L/opt/csw/lib -L/usr/local/lib&quot;
 </pre></div>
 
-<p>and for a 64-bit Sparc target
-</p><div class="example">
+<p>For an &lsquo;<samp>amd64</samp>&rsquo; target we used 
+</p>
+<div class="example">
 <pre class="example">CC=&quot;gcc -m64&quot;
+CPPFLAGS=&quot;-I/opt/csw/include -I/usr/local/include&quot;
 F77=&quot;gfortran -m64&quot;
+FPICFLAGS=-fPIC
 CXX=&quot;g++ -m64&quot;
 FC=$F77
-CPPFLAGS=-I/opt/csw/include
-LDFLAGS=&quot;-L/opt/csw/gcc4/lib/sparcv9 -L/opt/csw/lib/sparcv9&quot;
+FCPICFLAGS=$FPICFLAGS
+LDFLAGS=&quot;-L/opt/csw/lib/amd64 -L/usr/local/lib/amd64&quot;
 </pre></div>
 
-<p>Note that paths such as <samp>/opt/csw/gcc4/lib/sparcv9</samp> may need to
-be in the
+<p>Note that paths such as <samp>/opt/csw/lib</samp>, <samp>/usr/local/lib/amd64</samp>
+and <samp>/opt/csw/lib/amd64</samp> may need to be in the
 <a name="index-LD_005fLIBRARY_005fPATH-4"></a>
 <code>LD_LIBRARY_PATH</code> during configuration.
 </p>
-<p>The compilation can be tuned to a particular cpu: the <acronym>CRAN</acronym>
-check system uses <code>-mtune=niagara2</code>.
-</p>
-<p>Compilation for an &lsquo;<samp>x86</samp>&rsquo; target with <code>gcc</code>&nbsp;4.9.2
-needed
+
+<p>Compilation for a 32-bit Sparc target with <code>gcc</code>&nbsp;4.9.2 or
+<code>gcc</code>&nbsp;5.2.0 needed
 </p>
 <div class="example">
-<pre class="example">CC=&quot;/opt/csw/gcc4/bin/gcc -m32&quot;
+<pre class="example">CC=&quot;gcc -m32&quot;
+F77=&quot;gfortran -m32&quot;
+CXX=&quot;g++ -m32&quot;
+FC=$F77
 CPPFLAGS=&quot;-I/opt/csw/include -I/usr/local/include&quot;
-F77=&quot;/opt/csw/gcc4/bin/gfortran -m32&quot;
-CXX=&quot;/opt/csw/gcc4/bin/g++ -m32&quot;
-FC=&quot;/opt/csw/gcc4/bin/gfortran -m32&quot;
-LDFLAGS=&quot;-L/opt/csw/gcc4/lib -L/opt/csw/lib -L/usr/local/lib&quot;
+LDFLAGS=&quot;-L/opt/csw/lib -L/usr/local/lib&quot;
 </pre></div>
 
-<p>(<code>-L/opt/csw/lib</code> is needed since TeX Live was built using
-32-bit <code>gcc</code>, and we need <samp>/opt/csw/lib</samp> in
-<code>R_LD_LIBRARY_PATH</code>.)
-</p>
-<p>For an &lsquo;<samp>amd64</samp>&rsquo; target with <code>gcc</code>&nbsp;4.9.2
-we used 
-</p>
-<div class="example">
-<pre class="example">CC=&quot;/opt/csw/gcc4/bin/gcc -m64&quot;
-CPPFLAGS=&quot;-I/opt/csw/include -I/usr/local/include&quot;
-F77=&quot;/opt/csw/gcc4/bin/gfortran -m64&quot;
-FPICFLAGS=-fPIC
-CXX=&quot;/opt/csw/gcc4/bin/g++ -m64&quot;
+<p>and for a 64-bit Sparc target
+</p><div class="example">
+<pre class="example">CC=&quot;gcc -m64&quot;
+F77=&quot;gfortran -m64&quot;
+CXX=&quot;g++ -m64&quot;
 FC=$F77
-FCPICFLAGS=$FPICFLAGS
-LDFLAGS=&quot;-L/opt/csw/gcc4/lib/amd64 -L/opt/csw/lib/amd64&quot;
+CPPFLAGS=&quot;-I/opt/csw/include -I/usr/local/include&quot;
+LDFLAGS=&quot;-L/opt/csw/lib/sparcv9 -L/usr/local/lib/sparcv9&quot;
 </pre></div>
 
+<p>Note that paths such as <samp>/opt/csw/lib</samp>, <samp>/usr/local/lib/sparcv9</samp> or
+<samp>/opt/csw/lib/sparcv9</samp> may need to be in the
+<a name="index-LD_005fLIBRARY_005fPATH-5"></a>
+<code>LD_LIBRARY_PATH</code> during configuration.
+</p>
+<p>The compilation can be tuned to a particular cpu: the <acronym>CRAN</acronym>
+check system uses <code>-mtune=niagara2</code>.
+</p>
+
 <hr>
 <a name="AIX"></a>
 <a name="AIX-1"></a>
@@ -5559,8 +5595,7 @@
 <p>There have been few recent reports on FreeBSD.
 </p>
 <p>There is a &lsquo;port&rsquo; at <a href="https://www.freebsd.org/ports/math.html">https://www.freebsd.org/ports/math.html</a>, for
-R 3.0.2 at the time of writing.  Davor Cubranic reported some success
-on x86_64 FreeBSD 10.2 for R 3.2.2.
+R 3.3.1 at the time of writing.
 </p>
 <p>Use of ICU for collation and the <code>configure</code> option
 <samp>--with-internal-tzcode</samp> are desirable workarounds.
@@ -5574,7 +5609,9 @@
 <p>Ingo Feinerer installed R version 3.2.2 on OpenBSD 5.8 arch
 &lsquo;<samp>amd64</samp>&rsquo; (their name for &lsquo;<samp>x86_64</samp>&rsquo;).  Details of the build
 (and patches applied) are at
-<a href="http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/math/R/">http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/math/R/</a>.
+<a href="http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/math/R/">http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/math/R/</a>. (Downgrading
+the <code>zlib</code> requirement to 1.2.3 is against the advice of the R
+developers.)
 </p>
 <hr>
 <a name="Cygwin"></a>
@@ -5664,7 +5701,7 @@
 </p>
 <p>We have found that the build process for R is quite sensitive to
 the choice of tools: please follow our instructions <strong>exactly</strong>,
-even to the choice of particular versions of the tools.<a name="DOCF63" href="#FOOT63"><sup>63</sup></a>  The build process for add-on packages is somewhat more
+even to the choice of particular versions of the tools.<a name="DOCF69" href="#FOOT69"><sup>69</sup></a>  The build process for add-on packages is somewhat more
 forgiving, but we recommend using the exact toolset at first, and only
 substituting other tools once you are familiar with the process.
 </p>
@@ -5717,7 +5754,7 @@
 
 <p>It is essential that the directory containing the command line tools
 comes first or second in the path: there are typically like-named
-tools<a name="DOCF64" href="#FOOT64"><sup>64</sup></a> in other directories, and they will <strong>not</strong>
+tools<a name="DOCF70" href="#FOOT70"><sup>70</sup></a> in other directories, and they will <strong>not</strong>
 work. The ordering of the other directories is less important, but if in
 doubt, use the order above.
 </p>
@@ -5766,7 +5803,7 @@
 <a name="The-Inno-Setup-installer-1"></a>
 <h3 class="section">D.2 The Inno Setup installer</h3>
 
-<p>To make the installer package (<samp>R-3.3.1-win.exe</samp>) we
+<p>To make the installer package (<samp>R-3.3.1patched-win.exe</samp>) we
 currently require the Unicode version of Inno Setup 5.3.7 or later from
 <a href="http://jrsoftware.org/">http://jrsoftware.org/</a>. This is <em>not</em> included in
 <samp>Rtools*.exe</samp>.
@@ -5824,7 +5861,7 @@
 </p>
 <p>This toolchain uses <em>multilib</em>: that is there is a single front-end
 such as <code>gcc.exe</code> for each of the compilers and 32-bit (the
-default) and 64-bit compilation are selected by the flags<a name="DOCF65" href="#FOOT65"><sup>65</sup></a>  <samp>-m32</samp> and <samp>-m64</samp>
+default) and 64-bit compilation are selected by the flags<a name="DOCF71" href="#FOOT71"><sup>71</sup></a>  <samp>-m32</samp> and <samp>-m64</samp>
 respectively.  The tools are all 32-bit Windows executables and should
 be able to run on any current version of Windows&mdash;however you do need a
 64-bit version of Windows to build 64-bit R as the build process runs
@@ -6117,6 +6154,7 @@
 <tr><td></td><td valign="top"><a href="#index-LD_005fLIBRARY_005fPATH-2"><code>LD_LIBRARY_PATH</code></a>:</td><td>&nbsp;</td><td valign="top"><a href="#Using-FORTRAN">Using FORTRAN</a></td></tr>
 <tr><td></td><td valign="top"><a href="#index-LD_005fLIBRARY_005fPATH-3"><code>LD_LIBRARY_PATH</code></a>:</td><td>&nbsp;</td><td valign="top"><a href="#Compile-and-load-flags">Compile and load flags</a></td></tr>
 <tr><td></td><td valign="top"><a href="#index-LD_005fLIBRARY_005fPATH-4"><code>LD_LIBRARY_PATH</code></a>:</td><td>&nbsp;</td><td valign="top"><a href="#Using-gcc">Using gcc</a></td></tr>
+<tr><td></td><td valign="top"><a href="#index-LD_005fLIBRARY_005fPATH-5"><code>LD_LIBRARY_PATH</code></a>:</td><td>&nbsp;</td><td valign="top"><a href="#Using-gcc">Using gcc</a></td></tr>
 <tr><td></td><td valign="top"><a href="#index-LOCAL_005fSOFT"><code>LOCAL_SOFT</code></a>:</td><td>&nbsp;</td><td valign="top"><a href="#Windows-packages">Windows packages</a></td></tr>
 <tr><td colspan="4"> <hr></td></tr>
 <tr><th><a name="Environment-variable-index_en_letter-O">O</a></th><td></td><td></td></tr>
@@ -6233,205 +6271,224 @@
 <samp>-qopenmp</samp>.  This includes for <code>clang</code> 3.7.x and the
 Intel C compiler.</p>
 <h3><a name="FOOT12" href="#DOCF12">(12)</a></h3>
+<p>This is true for earlier versions of <code>g++</code> such
+as 4.2.1, and also for commonly-used versions of the Solaris compiler
+<code>CC</code>.</p>
+<h3><a name="FOOT13" href="#DOCF13">(13)</a></h3>
+<p>Currently only GCC 6 and
+later, but this has been mooted for others.</p>
+<h3><a name="FOOT14" href="#DOCF14">(14)</a></h3>
+<p>For when features were
+supported, see
+<a href="https://gcc.gnu.org/projects/cxx-status.html#cxx11">https://gcc.gnu.org/projects/cxx-status.html#cxx11</a>.</p>
+<h3><a name="FOOT15" href="#DOCF15">(15)</a></h3>
 <p>Suitable distributions include
 Strawberry Perl, <a href="http://strawberryperl.com/">http://strawberryperl.com/</a> and ActivePerl,
 <a href="https://www.activestate.com/activeperl">https://www.activestate.com/activeperl</a>.</p>
-<h3><a name="FOOT13" href="#DOCF13">(13)</a></h3>
+<h3><a name="FOOT16" href="#DOCF16">(16)</a></h3>
 <p>The installer as puts links to <code>R</code> and
 <code>Rscript</code> in <samp>/usr/bin</samp> (Mavericks, Yosemite) or
 <samp>/usr/local/bin</samp> (El Capitan and later). If these are missing, you
 can run directly the versions in
 <samp>/Library/Frameworks/R.framework/Resources/</samp>.</p>
-<h3><a name="FOOT14" href="#DOCF14">(14)</a></h3>
+<h3><a name="FOOT17" href="#DOCF17">(17)</a></h3>
 <p>unless they were excluded in the build.</p>
-<h3><a name="FOOT15" href="#DOCF15">(15)</a></h3>
+<h3><a name="FOOT18" href="#DOCF18">(18)</a></h3>
 <p>its binding is locked once the startup files have been
 read, so users cannot easily change it.</p>
-<h3><a name="FOOT16" href="#DOCF16">(16)</a></h3>
+<h3><a name="FOOT19" href="#DOCF19">(19)</a></h3>
 <p>If a proxy needs to be set, see
 <code>?download.file</code>.</p>
-<h3><a name="FOOT17" href="#DOCF17">(17)</a></h3>
+<h3><a name="FOOT20" href="#DOCF20">(20)</a></h3>
 <p>for a small number of
 <acronym>CRAN</acronym> packages where this is known to be safe and is needed by
 the autobuilder this is the default.  Look at the source of
 <samp>tools:::.install_packages</samp> for the list.  It can also be specified
 in the package&rsquo;s <samp>DESCRIPTION</samp> file.</p>
-<h3><a name="FOOT18" href="#DOCF18">(18)</a></h3>
+<h3><a name="FOOT21" href="#DOCF21">(21)</a></h3>
 <p>or by adding it in a file such as
 <samp>etc/i386/Makevars.site</samp>, which does not exist by default.</p>
-<h3><a name="FOOT19" href="#DOCF19">(19)</a></h3>
+<h3><a name="FOOT22" href="#DOCF22">(22)</a></h3>
 <p>They need to have been
 created using <samp>-headerpad_max_install_names</samp>, which is the
 default for an R package.</p>
-<h3><a name="FOOT20" href="#DOCF20">(20)</a></h3>
+<h3><a name="FOOT23" href="#DOCF23">(23)</a></h3>
 <p>&lsquo;X/Open Portability Guide&rsquo;, which has
 had several versions.</p>
-<h3><a name="FOOT21" href="#DOCF21">(21)</a></h3>
+<h3><a name="FOOT24" href="#DOCF24">(24)</a></h3>
 <p>On some systems setting
 <code>LC_ALL</code> or <code>LC_MESSAGES</code> to &lsquo;<samp>C</samp>&rsquo; disables <code>LANGUAGE</code>.</p>
-<h3><a name="FOOT22" href="#DOCF22">(22)</a></h3>
+<h3><a name="FOOT25" href="#DOCF25">(25)</a></h3>
 <p>If you try changing from French
 to Russian except in a UTF-8 locale, you will most likely find messages
 change to English.</p>
-<h3><a name="FOOT23" href="#DOCF23">(23)</a></h3>
+<h3><a name="FOOT26" href="#DOCF26">(26)</a></h3>
 <p>the
 language written in England: some people living in the USA appropriate
 this name for their language.</p>
-<h3><a name="FOOT24" href="#DOCF24">(24)</a></h3>
+<h3><a name="FOOT27" href="#DOCF27">(27)</a></h3>
 <p>with
 Americanisms.</p>
-<h3><a name="FOOT25" href="#DOCF25">(25)</a></h3>
+<h3><a name="FOOT28" href="#DOCF28">(28)</a></h3>
 <p>also known as
 <acronym>IEEE</acronym>&nbsp;754</p>
-<h3><a name="FOOT26" href="#DOCF26">(26)</a></h3>
+<h3><a name="FOOT29" href="#DOCF29">(29)</a></h3>
 <p>at least when storing quantities: the on-FPU
 precision is allowed to vary</p>
-<h3><a name="FOOT27" href="#DOCF27">(27)</a></h3>
+<h3><a name="FOOT30" href="#DOCF30">(30)</a></h3>
 <p>e.g. Bessel, beta and gamma functions</p>
-<h3><a name="FOOT28" href="#DOCF28">(28)</a></h3>
+<h3><a name="FOOT31" href="#DOCF31">(31)</a></h3>
 <p>including copying <samp>MkRules.dist</samp> to
 <samp>MkRule.local</samp> and selecting the architecture.</p>
-<h3><a name="FOOT29" href="#DOCF29">(29)</a></h3>
+<h3><a name="FOOT32" href="#DOCF32">(32)</a></h3>
 <p>also known as
 <acronym>IEEE</acronym>&nbsp;754</p>
-<h3><a name="FOOT30" href="#DOCF30">(30)</a></h3>
+<h3><a name="FOOT33" href="#DOCF33">(33)</a></h3>
 <p>Note
 that C11 compilers need not be C99-compliant: R requires support for
 <code>double complex</code> and variable-length arrays which are optional in
 C11 but is mandatory in C99.</p>
-<h3><a name="FOOT31" href="#DOCF31">(31)</a></h3>
+<h3><a name="FOOT34" href="#DOCF34">(34)</a></h3>
 <p><samp>-std=c99</samp> excludes POSIX
 functionality, but <samp>config.h</samp> will turn on all <acronym>GNU</acronym>
 extensions to include the POSIX functionality.  The default mode for GCC
 5.1 and later is <samp>-std=gnu11</samp>.</p>
-<h3><a name="FOOT32" href="#DOCF32">(32)</a></h3>
+<h3><a name="FOOT35" href="#DOCF35">(35)</a></h3>
 <p>However, it is possible to break
 the default behaviour of <code>glibc</code> by re-specifying the <code>gconv</code>
 modules to be loaded.</p>
-<h3><a name="FOOT33" href="#DOCF33">(33)</a></h3>
+<h3><a name="FOOT36" href="#DOCF36">(36)</a></h3>
 <p>specifically, the C99
 functionality of headers <samp>wchar.h</samp> and <samp>wctype.h</samp>, types
 <code>wctans_t</code> and <code>mbstate_t</code> and functions <code>mbrtowc</code>,
 <code>mbstowcs</code>, <code>wcrtomb</code>, <code>wcscoll</code>, <code>wcstombs</code>,
 <code>wctrans</code>, <code>wctype</code>, and <code>iswctype</code>.</p>
-<h3><a name="FOOT34" href="#DOCF34">(34)</a></h3>
+<h3><a name="FOOT37" href="#DOCF37">(37)</a></h3>
 <p>including the <code>opendir</code>, <code>readdir</code>,
 <code>closedir</code>, <code>popen</code>, <code>stat</code>, <code>glob</code>, <code>access</code>,
 <code>getcwd</code> and <code>chdir</code> system calls, and either <code>putenv</code> or
 <code>setenv</code>.</p>
-<h3><a name="FOOT35" href="#DOCF35">(35)</a></h3>
+<h3><a name="FOOT38" href="#DOCF38">(38)</a></h3>
 <p>such as
 <code>realpath</code>, <code>symlink</code>.</p>
-<h3><a name="FOOT36" href="#DOCF36">(36)</a></h3>
+<h3><a name="FOOT39" href="#DOCF39">(39)</a></h3>
 <p>most often distributed as part of <code>xz</code>:
 possible names in Linux distributions include
 <code>xz-devel</code>/<code>xz-libs</code> and <code>liblzma-dev</code>.</p>
-<h3><a name="FOOT37" href="#DOCF37">(37)</a></h3>
+<h3><a name="FOOT40" href="#DOCF40">(40)</a></h3>
 <p>sometimes known as PCRE1, and not PCRE2 which started at
 version 10.0.</p>
-<h3><a name="FOOT38" href="#DOCF38">(38)</a></h3>
-<p>for
-example to specify static linking with a build which has both shared and
-static libraries.</p>
-<h3><a name="FOOT39" href="#DOCF39">(39)</a></h3>
+<h3><a name="FOOT41" href="#DOCF41">(41)</a></h3>
+<p>but not a major
+version greater than 7 should there ever be one: the major version has
+been 7 since 2000.</p>
+<h3><a name="FOOT42" href="#DOCF42">(42)</a></h3>
+<p>for example to specify static linking with a
+build which has both shared and static libraries.</p>
+<h3><a name="FOOT43" href="#DOCF43">(43)</a></h3>
 <p>Such as
 <acronym>GNU</acronym> <code>tar</code> 1.15 or later, <code>bsdtar</code> (from
 <a href="https://github.com/libarchive/libarchive/">https://github.com/libarchive/libarchive/</a>, as used by FreeBSD and OS
 X 10.6 and later) or <code>tar</code> from the Heirloom Toolchest
 (<a href="http://heirloom.sourceforge.net/tools.html">http://heirloom.sourceforge.net/tools.html</a>).</p>
-<h3><a name="FOOT40" href="#DOCF40">(40)</a></h3>
+<h3><a name="FOOT44" href="#DOCF44">(44)</a></h3>
 <p><code>texi2dvi</code> is normally a shell
 script.  Some versions (including that from <strong>texinfo</strong> 5.2 and 6.0)
 need to be run under <code>bash</code> rather than a Bourne shell,
 especially on Solaris. Some of the issues which have been observed with
 broken versions of <code>texi2dvi</code> can be circumvented by setting the
 environment variable <code>R_TEXI2DVICMD</code> to the value <code>emulation</code>.</p>
-<h3><a name="FOOT41" href="#DOCF41">(41)</a></h3>
+<h3><a name="FOOT45" href="#DOCF45">(45)</a></h3>
 <p>If necessary the path to
 <code>pkg-config</code> can be specified by setting <code>PKGCONF</code> in
 <samp>config.site</samp>, on the <code>configure</code> command line or in the
 environment.</p>
-<h3><a name="FOOT42" href="#DOCF42">(42)</a></h3>
+<h3><a name="FOOT46" href="#DOCF46">(46)</a></h3>
 <p>also known as <code>ttf-mscorefonts-installer</code> in the
 Debian/Ubuntu world: see also
 <a href="https://en.wikipedia.org/wiki/Core_fonts_for_the_Web">https://en.wikipedia.org/wiki/Core_fonts_for_the_Web</a>.</p>
-<h3><a name="FOOT43" href="#DOCF43">(43)</a></h3>
+<h3><a name="FOOT47" href="#DOCF47">(47)</a></h3>
 <p><code>ttf-liberation</code>
 in Debian/Ubuntu.</p>
-<h3><a name="FOOT44" href="#DOCF44">(44)</a></h3>
+<h3><a name="FOOT48" href="#DOCF48">(48)</a></h3>
 <p>This is true even for
 the &lsquo;Aqua&rsquo; version of Tk on OS X, but distributions of that include a
 copy of the X11 files needed.</p>
-<h3><a name="FOOT45" href="#DOCF45">(45)</a></h3>
-<p>Using the Oracle Solaris Studio
+<h3><a name="FOOT49" href="#DOCF49">(49)</a></h3>
+<p>Using the Oracle Developer Studio
 <code>cc</code> and <code>f95</code> compilers</p>
-<h3><a name="FOOT46" href="#DOCF46">(46)</a></h3>
+<h3><a name="FOOT50" href="#DOCF50">(50)</a></h3>
 <p>and &lsquo;<samp>i686</samp>&rsquo; for earlier
 versions.</p>
-<h3><a name="FOOT47" href="#DOCF47">(47)</a></h3>
+<h3><a name="FOOT51" href="#DOCF51">(51)</a></h3>
 <p>We have measured 15&ndash;20% on &lsquo;<samp>i686</samp>&rsquo; Linux
 and around 10% on &lsquo;<samp>x86_64</samp>&rsquo; Linux.</p>
-<h3><a name="FOOT48" href="#DOCF48">(48)</a></h3>
+<h3><a name="FOOT52" href="#DOCF52">(52)</a></h3>
 <p>We believe that
 versions 3.4.0 to 3.10.1 are compatible.</p>
-<h3><a name="FOOT49" href="#DOCF49">(49)</a></h3>
+<h3><a name="FOOT53" href="#DOCF53">(53)</a></h3>
 <p>On HP-UX <code>fort77</code> is the
 <acronym>POSIX</acronym> compliant FORTRAN compiler, and comes after
 <code>g77</code>.</p>
-<h3><a name="FOOT50" href="#DOCF50">(50)</a></h3>
+<h3><a name="FOOT54" href="#DOCF54">(54)</a></h3>
 <p>as well as its equivalence to the <code>Rcomplex</code>
 structure defined in <samp>R_ext/Complex.h</samp>.</p>
-<h3><a name="FOOT51" href="#DOCF51">(51)</a></h3>
+<h3><a name="FOOT55" href="#DOCF55">(55)</a></h3>
 <p>for example, <code>X11 font at size 14 could not
 be loaded</code>.</p>
-<h3><a name="FOOT52" href="#DOCF52">(52)</a></h3>
+<h3><a name="FOOT56" href="#DOCF56">(56)</a></h3>
 <p>or <code>-mtune=corei7</code> for Intel Core
 i3/15/17 with <code>gcc &gt;= 4.6.0</code>.</p>
-<h3><a name="FOOT53" href="#DOCF53">(53)</a></h3>
+<h3><a name="FOOT57" href="#DOCF57">(57)</a></h3>
 <p>This also needs the OpenMP runtime,
-which is usually distributed separately, e.g. at
+which has sometimes been distributed separately, e.g. for 3.7.0 at
 <a href="http://llvm.org/releases">http://llvm.org/releases</a>.</p>
-<h3><a name="FOOT54" href="#DOCF54">(54)</a></h3>
+<h3><a name="FOOT58" href="#DOCF58">(58)</a></h3>
 <p>It will be necessary to
 install later versions of software such as <code>libcurl</code>.</p>
-<h3><a name="FOOT55" href="#DOCF55">(55)</a></h3>
+<h3><a name="FOOT59" href="#DOCF59">(59)</a></h3>
 <p>Apple provides a partial emulation of
 GNU readline 4.2 based on the NetBSD <code>editline</code> library.  That is
 not recommended but for the time being R&rsquo;s installation scripts will
-make use of it.</p>
-<h3><a name="FOOT56" href="#DOCF56">(56)</a></h3>
+make use of it if GNU readline is not found.</p>
+<h3><a name="FOOT60" href="#DOCF60">(60)</a></h3>
 <p>These
 days that is defined by Apple&rsquo;s implementation of <code>clang</code>, so it is
 strongly recommended to use that.</p>
-<h3><a name="FOOT57" href="#DOCF57">(57)</a></h3>
+<h3><a name="FOOT61" href="#DOCF61">(61)</a></h3>
 <p>This
 is a tarball which needs to be unpacked in the Terminal by e.g.
 <code>sudo tar -zxf gfortran-4.8.2-darwin13.tar.bz2 -C /</code>. It does
 not run on Core 2 Duo Macs.</p>
-<h3><a name="FOOT58" href="#DOCF58">(58)</a></h3>
+<h3><a name="FOOT62" href="#DOCF62">(62)</a></h3>
 <p>This also needs the OpenMP runtime,
-which is distributed separately at that site.</p>
-<h3><a name="FOOT59" href="#DOCF59">(59)</a></h3>
+which was distributed separately at that site for 3.7.0.</p>
+<h3><a name="FOOT63" href="#DOCF63">(63)</a></h3>
 <p>It is reported that for some non-Apple
 toolchains <code>CPPFLAGS</code> needed to contain <code>-D__ACCELERATE__</code>.</p>
-<h3><a name="FOOT60" href="#DOCF60">(60)</a></h3>
+<h3><a name="FOOT64" href="#DOCF64">(64)</a></h3>
+<p>E.g. <em>via</em>
+<code>tlmgr install cm-super helvetic inconsolata texinfo</code> .</p>
+<h3><a name="FOOT65" href="#DOCF65">(65)</a></h3>
 <p>For more
 details see <a href="http://www.macstrategy.com/article.php?3">http://www.macstrategy.com/article.php?3</a>.</p>
-<h3><a name="FOOT61" href="#DOCF61">(61)</a></h3>
+<h3><a name="FOOT66" href="#DOCF66">(66)</a></h3>
 <p>e.g.
 <code>Java For OS X 2015-001</code> from
 <a href="https://support.apple.com/kb/DL1572">https://support.apple.com/kb/DL1572</a>.</p>
-<h3><a name="FOOT62" href="#DOCF62">(62)</a></h3>
-<p>including <code>gcc</code> for
-Sparc from Oracle.</p>
-<h3><a name="FOOT63" href="#DOCF63">(63)</a></h3>
+<h3><a name="FOOT67" href="#DOCF67">(67)</a></h3>
+<p>Oracle Solaris Studio prior to 2016, and previously Sun
+Studio.</p>
+<h3><a name="FOOT68" href="#DOCF68">(68)</a></h3>
+<p>using the <samp>-V</samp> flag.</p>
+<h3><a name="FOOT69" href="#DOCF69">(69)</a></h3>
 <p>For
 example, the Cygwin version of <code>make 3.81</code> fails to work
 correctly.</p>
-<h3><a name="FOOT64" href="#DOCF64">(64)</a></h3>
+<h3><a name="FOOT70" href="#DOCF70">(70)</a></h3>
 <p>such as <code>sort</code>, <code>find</code> and perhaps
 <code>make</code>.</p>
-<h3><a name="FOOT65" href="#DOCF65">(65)</a></h3>
+<h3><a name="FOOT71" href="#DOCF71">(71)</a></h3>
 <p>these
 flags apply to the compilers: some of the tools use different flags.
 32-bit builds are the default.</p>
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/manual/R-FAQ.texi R-patched/doc/manual/R-FAQ.texi
--- R-release-orig/doc/manual/R-FAQ.texi	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/doc/manual/R-FAQ.texi	2016-07-24 17:45:10.716574000 +0200
@@ -283,11 +283,11 @@
 
 The current version of R will configure and build under a number of
 common Unix-like (e.g., @uref{https://en.wikipedia.org/wiki/Unix-like})
-platforms including @var{cpu}-linux-gnu for the i386, amd64, alpha,
-arm/armel, hppa, ia64, m68k, mips/mipsel, powerpc, s390 and sparc 
-@acronym{CPU}s (e.g., @url{https://buildd.debian.org/build.php?&pkg=r-base}),
-i386-hurd-gnu, @var{cpu}-kfreebsd-gnu for i386 and amd64, 
-powerpc-apple-darwin, mips-sgi-irix,
+platforms including @var{cpu}-linux-gnu for the i386, amd64/x86_64,
+alpha, arm, arm64, hppa, mips/mipsel, powerpc, s390x and sparc
+@acronym{CPU}s (e.g.,
+@url{https://buildd.debian.org/build.php?&pkg=r-base}), i386-hurd-gnu,
+@var{cpu}-kfreebsd-gnu for i386 and amd64,
 @c <FIXME>
 @c Not sure anymore ...
 @c alpha-dec-osf4,
@@ -295,12 +295,14 @@
 @c i386-freebsd reported by B. Gian James" <gian.james@gmail.com> on
 @c 2009-01-11
 @c (Architecture: i386, OS: FreeBSD 7.1-PRERELEASE, Vendor: PC-BSD)
-i386-freebsd,
+@c i386-freebsd,
+@c but all recent reports are on x86_64
 @c </COMMENT>
-@c i386-sun-solaris,
 @c hppa-hp-hpux, 
+@c mips-sgi-irix,
 @c </FIXME>
-rs6000-ibm-aix, and sparc-sun-solaris.
+i386-pc-solaris, rs6000-ibm-aix, sparc-sun-solaris, x86_64-apple-darwin,
+x86_64-unknown-freebsd and x86_64-unknown-openbsd.
 
 @c and according to @email{jlindsey@@luc.ac.be, Jim Lindsey} also on
 @c Mac, Amiga and Atari under m68k-linux.
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/manual/R-admin.texi R-patched/doc/manual/R-admin.texi
--- R-release-orig/doc/manual/R-admin.texi	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/doc/manual/R-admin.texi	2016-07-27 02:41:07.820420000 +0200
@@ -937,9 +937,10 @@
 
 @menu
 * OpenMP Support::              
+* C++ Support::                 
 @end menu
 
-@node OpenMP Support,  , Other Options, Other Options
+@node OpenMP Support, C++ Support, Other Options, Other Options
 @subsection OpenMP Support
 
 By default @command{configure} searches for suitable
@@ -974,6 +975,64 @@
 object and loading it into the C program of @R{}'s executable.  Note
 that overridden values are not tested.
 
+@node C++ Support,  , OpenMP Support, Other Options
+@subsection C++ Support
+
+C++ is not used by @R{} itself, but support is provided for installing
+packages with C++ code via @command{make} macros defined in file
+@file{etc/Makeconf} (and with explanations in file @file{config.site}):
+@example
+CXX
+CXXFLAGS
+CXXPICFLAGS
+
+CXX1X
+CXX1XSTD
+CXX1XFLAGS
+CXX1XPICFLAGS
+@end example
+@noindent
+The macros @code{CXX} etc are those used by default for C++ code: their
+values are not checked.  @command{configure} will attempt to set the
+rest suitably, choosing for @code{CXX1XSTD} a suitable flag such as
+@option{-std=c++11} for C++11 support.  The inferred values can be
+overridden in file @file{config.site} or on the @command{configure}
+command line: user-supplied values will be tested compiling some C++11
+code.
+
+It may be@footnote{This is true for earlier versions of @command{g++} such
+as 4.2.1, and also for commonly-used versions of the Solaris compiler
+@code{CC}.} that there is no suitable flag for C++11 support, in which
+case a different compiler could be selected for @code{CXX1X} and its
+corresponding flags.  Some compilers@footnote{Currently only GCC 6 and
+later, but this has been mooted for others.} by default assume a later
+standard than C++98 whereas the latter is assumed by some packages.  So
+users of GCC 6 might like to specify
+@example
+CXX='g++ -std=gnu++98'
+CXX1X=g++
+CXX1XSTD='-std=c++11'
+@end example
+
+The @option{-std} flag is supported by the GCC, @command{clang}, Intel
+and Solaris compilers (the latter from version 12.4).  Currently
+accepted values are (plus some synonyms)
+@example
+g++:     c++98 gnu++98 c++11 gnu+11 c++14 gnu++14 c++1z gnu++1z
+Intel:   gnu+98 c++11 c++14 (from 16.0)
+Solaris: c++03 c++11 c++14 (from 12.5)
+@end example
+@noindent
+(Those for @command{clang++} are not documented, but seem to be based on
+@code{g++}.)  Versions 4.3.x to 4.8.x of @command{g++} accepted flag
+@option{-std=c++0x} with partial support@footnote{For when features were
+supported, see
+@uref{https://gcc.gnu.org/projects/cxx-status.html#cxx11}.} for C++11:
+this is currently still accepted as a deprecated synonym for
+@option{-std=c++11}.
+
+For the use of C++11 in @R{} packages see the `Writing R Extensions' manual.
+
 @node Testing a Unix-alike Installation,  , Other Options, Installing R under Unix-alikes
 @section Testing an Installation
 
@@ -1350,7 +1409,7 @@
 @node Support for libcurl, Checking the build, Using ICU for collation, Building from source
 @subsection Support for libcurl
 
-@code{libcurl} version 7.28.0 or later can be used to support
+@code{libcurl} version 7.28.0 or later is used to support
 @code{curlGetHeaders} and the @code{"libcurl"} methods of
 @code{download.file} and @code{url}.
 
@@ -1645,6 +1704,9 @@
 @sc{R.app}
 @end macro
 
+(`OS X' will be renamed to `macOS' as from version 10.12 `macOS Sierra'
+in late 2016.)
+
 The front page of a @acronym{CRAN} site has a link `Download R for OS
 X'. Click on that, then download the file @file{R-@value{VERSIONno}.pkg}
 and install it. This runs on OS X 10.9 and later (Mavericks, Yosemite,
@@ -3174,10 +3236,8 @@
 @code{setenv}.} are essential, and others@footnote{such as
 @code{realpath}, @code{symlink}.} will be used if available.
 
-@c PCRE[1] 8.32 is from Nov 2012
 @c zlib 1.2.5 is from July 2010, bzip2 1.0.6 from Sept 2010
 @c xz 5.0.3 is from May 2011
-@c libcurl 7.28.0 was released in Oct 2012
 Installations of @code{zlib} (version 1.2.5 or later), @code{libbz2}
 (version 1.0.6 or later: called @pkg{bzip2-libs}/@pkg{bzip2-devel} or
 @pkg{libbz2-1.0}/@pkg{libbz2-dev} by some Linux distributions),
@@ -3186,6 +3246,7 @@
 @code{xz-devel}/@code{xz-libs} and @code{liblzma-dev}.} version 5.0.3 or
 later are required.
 
+@c PCRE[1] 8.32 is from Nov 2012, but Ubuntu 14.04LTS does not have it
 PCRE@footnote{sometimes known as PCRE1, and not PCRE2 which started at
 version 10.0.} (version 8.32 or later, although versions 8.10--8.31 will
 be accepted with a deprecation warning) is required (or just its library
@@ -3202,12 +3263,14 @@
 @noindent  
 The @option{--enable-jit} flag is supported for most common CPUs.
 
-Library @code{libcurl} (version 7.28.0 or later) is required.
-Information on @code{libcurl} is found from the @command{curl-config}
-script: if that is missing or needs to be overridden@footnote{for
-example to specify static linking with a build which has both shared and
-static libraries.} there are macros to do so described in file
-@file{config.site}.
+@c libcurl 7.28.0 was released in Oct 2012
+Library @code{libcurl} (version 7.28.0 or later@footnote{but not a major
+version greater than 7 should there ever be one: the major version has
+been 7 since 2000.}) is required.  Information on @code{libcurl} is
+found from the @command{curl-config} script: if that is missing or needs
+to be overridden@footnote{for example to specify static linking with a
+build which has both shared and static libraries.} there are macros to
+do so described in file @file{config.site}.
 
 A @command{tar} program is needed to unpack the sources and packages
 (including the recommended packages).  A version@footnote{Such as
@@ -3273,6 +3336,15 @@
 The essential programs should be in your @env{PATH} at the time
 @command{configure} is run: this will capture the full paths.
 
+Those distributing binary versions of @R{} may need to be aware of the
+licences of the external libraries it is linked to (including `useful'
+libraries from the next section).  The @code{liblzma} library is in the
+public domain and X11, @code{libbzip2}, @code{libcurl} and @code{zlib}
+have MIT-style licences. PCRE has a BSD-style licence which requires
+distribution of the licence (included in @R{}'s @file{COPYRIGHTS} file)
+in binary distributions.  GNU @code{readline} is licensed under GPL
+(which version(s) depending on the @code{readline} version).
+
 @node Useful libraries and programs, Linear algebra, Essential programs and libraries, Essential and useful other programs under a Unix-alike
 @section Useful libraries and programs
 
@@ -3340,14 +3412,6 @@
 at @url{https://github.com/laurikari/tre/}, but at the time of writing
 the resulting build will not pass its checks.).
 
-@c 5.0.3 was released on 2011-05-21
-Library @code{liblzma} from @code{xz-utils} version 5.0.3 or later
-(including 5.2.x) will be used if installed: the version in the @R{}
-sources can be selected instead by configuring with
-@option{--without-system-xz}.  Systems differ in what they call the
-package including this: e.g.@: on Fedora the library is in
-@samp{xz-libs} and the headers in @samp{xz-devel}.
-
 An implementation of @acronym{XDR} is required, and the @R{} sources
 contain one which is likely to suffice (although a system version may
 have higher performance).  @acronym{XDR} is part of @acronym{RPC} and
@@ -3482,10 +3546,10 @@
 most cases this will be a C++98 compiler, and as from @R{} 3.1.0 it is
 possible to specify an alternative compiler for use with C++11 by the
 configure variables @code{CXX1X}, @code{CXX1XSTD}, @code{CXX1XFLAGS} and
-similar.  Again, @command{configure} will normally find a suitable value
-for @code{CXX1XSTD} if the compiler given by @code{CXX} is capable of
-compiling C++11 code, but it is possible that a completely different
-compiler will be needed.
+similar (@pxref{C++ Support}).  Again, @command{configure} will normally
+find a suitable value for @code{CXX1XSTD} if the compiler given by
+@code{CXX} is capable of compiling C++11 code, but it is possible that a
+completely different compiler will be needed.
 
 Other packages need full Fortran 90 (or later) support.  For source
 files with extension @file{.f90} or @file{.f95}, the compiler defined by
@@ -3547,7 +3611,7 @@
 position-independent code), and that is not checked.
 
 Some enhanced @acronym{BLAS}es are compiler-system-specific
-(@code{sunperf} on Solaris@footnote{Using the Oracle Solaris Studio
+(@code{sunperf} on Solaris@footnote{Using the Oracle Developer Studio
 @command{cc} and @command{f95} compilers}, @code{libessl} on IBM,
 @code{Accelerate} on OS X).  The correct incantation for these is often
 found @emph{via} @option{--with-blas} with no value on the appropriate
@@ -4597,7 +4661,6 @@
 @menu
 * Clang::                       
 * Intel compilers::             
-* Oracle Solaris Studio compilers::  
 @end menu
 
 @node Clang, Intel compilers, Linux, Linux
@@ -4620,10 +4683,10 @@
 
 Most builds of @command{clang} have no OpenMP support.  Builds of
 version 3.7.0 or later may.@footnote{This also needs the OpenMP runtime,
-which is usually distributed separately, e.g.@: at
+which has sometimes been distributed separately, e.g.@: for 3.7.0 at
 @uref{http://llvm.org/releases}.}
 
-@node Intel compilers, Oracle Solaris Studio compilers, Clang, Linux
+@node Intel compilers,  , Clang, Linux
 @subsection Intel compilers
 
 Intel compilers have been used under @cputype{ix86} and @cputype{x86_64}
@@ -4675,47 +4738,13 @@
 FCFLAGS=$F77FLAGS
 @end example
 
-
-@node Oracle Solaris Studio compilers,  , Intel compilers, Linux
-@subsection Oracle Solaris Studio compilers
-
-Brian Ripley tested the Sun Studio 12 compilers, since renamed to Oracle
-Solaris Studio.  On @cputype{x86_64} Linux with
-
-@example
-CC=suncc
-CFLAGS="-xO5 -xc99 -xlibmil -nofstore"
-CPICFLAGS=-Kpic
-F77=sunf95
-FFLAGS="-O5 -libmil -nofstore"
-FPICFLAGS=-Kpic
-CXX="sunCC -library=stlport4"
-CXXFLAGS="-xO5 -xlibmil -nofstore -features=tmplrefstatic"
-CXXPICFLAGS=-Kpic
-FC=sunf95
-FCFLAGS=$FFLAGS
-FCPICFLAGS=-Kpic
-LDFLAGS=-L/opt/sunstudio12.1/rtlibs/amd64
-SHLIB_LDFLAGS=-shared
-SHLIB_CXXLDFLAGS=-G
-SHLIB_FCLDFLAGS=-G
-SAFE_FFLAGS="-O5 -libmil"
-@end example
-
-@noindent
-@option{-m64} could be added, but was the default.  Do not use
-@option{-fast}: see the warnings under Solaris. (The C++ options are
-also explained under Solaris.)
-
-Others have found on at least some versions of @cputype{ix86} Linux that
-the configure flag @option{--disable-largefile} was needed (since
-@file{glob.h} on that platform presumed @command{gcc} was being used).
-
-
 @node OS X, Solaris, Linux, Platform notes
 @section OS X
 @cindex OS X
 
+(`OS X' will be renamed to `macOS' from version 10.12 `macOS Sierra' in
+late 2016.)
+
 The instructions here are for @cputype{x86_64} builds on 10.9
 (Mavericks) or later.  In principle@footnote{It will be necessary to
 install later versions of software such as @code{libcurl}.} @R{} can be
@@ -4731,10 +4760,11 @@
 You need @code{readline}@footnote{Apple provides a partial emulation of
 GNU readline 4.2 based on the NetBSD @code{editline} library.  That is
 not recommended but for the time being @R{}'s installation scripts will
-make use of it.} and a Fortran compiler.  Those and other binary
-components are available from @uref{https://r.research.att.com/libs}:
-you are likely to need @code{pcre} and @code{xz} (recent OS X provides
-libraries but not headers for these).
+make use of it if GNU readline is not found.} and a Fortran compiler.
+Those and other binary components are available from
+@uref{https://r.research.att.com/libs}: you are likely to need
+@code{pcre} and @code{xz} (recent OS X provides libraries but not
+headers for these).
 
 An X sub-system is required unless configuring using
 @option{--without-x}: see @uref{https://xquartz.macosforge.org/}.
@@ -4788,7 +4818,8 @@
 available from @uref{http://llvm.org/releases/}.  In particular, these
 include support for the `Address Sanitizer' (not included by Apple until
 Xcode 7) and for OpenMP@footnote{This also needs the OpenMP runtime,
-which is distributed separately at that site.} in version 3.7.0 and later.
+which was distributed separately at that site for 3.7.0.} in version
+3.7.0 and later.
 
 Pre-compiled versions of many of the @ref{Useful libraries and programs}
 are available from @uref{https://r.research.att.com/libs/}.  You will
@@ -4842,9 +4873,11 @@
 
 The @TeX{} implementation used by the developers is MacTeX
 (@uref{https://www.tug.org/mactex/}): the full installation is about
-4GB, but a smaller version is available at
+5GB, but a smaller version is available at
 @uref{https://www.tug.org/mactex/morepackages.html}: you will need to
-add some packages, e.g.@: for the 2015 version we needed to add
+add some packages, e.g.@: for the 2016 version we needed to
+add@footnote{E.g.@: @emph{via}
+@command{tlmgr install cm-super helvetic inconsolata texinfo} .}
 @pkg{cm-super}, @pkg{helvetic}, @pkg{inconsolata} and @pkg{texinfo}
 which brought this to about 410MB.  @samp{TeX Live Utility} (available
 @emph{via} the MacTeX front page) provides a graphical means to manage
@@ -4872,14 +4905,15 @@
 @node El Capitan, Tcl/Tk headers and libraries, OS X, OS X
 @subsection El Capitan
 
-There are problems resulting from the new-to-El-Capitan restriction that
+There have been problems resulting from the new-to-El-Capitan restriction that
 only Apple is allowed to install software under @file{/usr}: this
-affects @emph{inter alia} @code{MacTeX} and @code{XQuartz}.  For
-@code{MacTeX} it is necessary to include @file{/Library/TeX/texbin} in
-your path rather than @file{/usr/texbin}.  Upgrading will move
-disallowed files to under @file{/Library/SystemMigration/usr}: this
-includes @file{/usr/X11R6}, @file{/usr/texbin}, @file{/usr/bin/R},
-@file{/usr/bin/Rscript} but not the link @file{/usr/X11}.
+affects @emph{inter alia} @code{MacTeX} and @code{XQuartz} which changed
+their installation locations.  For @code{MacTeX} it is necessary to
+include @file{/Library/TeX/texbin} in your path rather than
+@file{/usr/texbin}.  Upgrading will move disallowed files to under
+@file{/Library/SystemMigration/usr}: this includes @file{/usr/X11R6},
+@file{/usr/texbin}, @file{/usr/bin/R}, @file{/usr/bin/Rscript} but not
+the link @file{/usr/X11}.
 
 @command{configure} can be told to look for @code{X11} in
 @code{XQuartz}'s main location of @file{/opt/X11}, e.g.@: by
@@ -4888,8 +4922,7 @@
 @end example
 
 @noindent
-although the linked versions under @file{/usr/X11} will be found (if the
-link is present).
+although linked versions under @file{/usr/X11} will be found.
 
 @node Tcl/Tk headers and libraries, Java (OS X), El Capitan, OS X
 @subsection Tcl/Tk headers and libraries
@@ -5052,15 +5085,18 @@
 @cindex Solaris
 
 @R{} has been built successfully on Solaris 10 (both Sparc and
-@cputype{x86}) using the (zero cost) Oracle Solaris Studio compilers:
-there has been some success with
-@command{gcc}@tie{}4/@command{gfortran}.  (Recent Sun machines are AMD
-Opterons or Intel Xeons (@cputype{amd64}) rather than @cputype{x86}, but
-32-bit @cputype{x86} executables are the default.)
+@cputype{x86}) using the (zero cost) Oracle Developer
+Studio@footnote{Oracle Solaris Studio prior to 2016, and previously Sun
+Studio.} compilers: there has been some success with
+@command{gcc}/@command{gfortran}.  (Recent Sun machines are AMD Opterons
+or Intel Xeons (@cputype{amd64}) rather than @cputype{x86}, but 32-bit
+@cputype{x86} executables are the default.)  How these compilers
+identify@footnote{using the @option{-V} flag.} themselves is slightly
+confusing: Solaris Studio versions 12.3 and 12.4 report versions 5.12
+and 5.13, and Developer Studio 12.5 reports 5.14.  We will only consider
+12.3 (Nov 2011) and later.
 
 There have been few reports on Solaris 11, with no known extra issues.
-Solaris 9 and earlier are now so old that it is unlikely that @R{} is
-still used with them, and they will not be considered here.
 
 The Solaris versions of several of the tools needed to build @R{}
 (e.g.@: @command{make}, @command{ar} and @command{ld}) are in
@@ -5103,9 +5139,9 @@
 you find time-zone abbreviations being given odd values (as has been
 seen on 64-bit builds without it).
 
-When using the Oracle compilers@footnote{including @command{gcc} for
-Sparc from Oracle.} do @emph{not} specify @option{-fast}, as this
-disables @acronym{IEEE} arithmetic and @command{make check} will fail.
+When using the Oracle compilers do @emph{not} specify @option{-fast}, as
+this disables @acronym{IEEE} arithmetic and @command{make check} will
+fail.
 
 It has been reported that some Solaris installations need
 
@@ -5128,18 +5164,21 @@
 FFLAGS=-O
 CXX="CC -library=stlport4"
 CXXFLAGS=-O
-FC=f95
+FC=$F77
 FCFLAGS=$FFLAGS
-FCLIBS="-lfai -lfsu"
-R_LD_LIBRARY_PATH="/usr/local/lib:/opt/csw/gcc4/lib:/opt/csw/lib"
+FCLIBS="-lfai -lfui -lfsu"
+R_LD_LIBRARY_PATH="/usr/local/lib:/opt/csw/lib"
 @end example
 
 @noindent
+(For version 12.5 of the compiler use @code{FCLIBS=-lfsu}.)
+
 For a 64-bit target add @option{-m64} to the compiler macros
-and use something like @code{LDFLAGS=-L/usr/local/lib/sparcv9} or
-@code{LDFLAGS=-L/usr/local/lib/amd64} as appropriate.
+and use something like @code{LDFLAGS=-L/usr/local/lib/amd64} or
+@code{LDFLAGS=-L/usr/local/lib/sparcv9} as appropriate (and other 64-bit
+library directories if used, e.g.@: @code{-L/opt/csw/lib/arm64}).
 It will also be necessary to point @command{pkg-config} at the 64-bit
-directories, e.g.@: one of
+directories, e.g.@: something like one of
 
 @example
 PKG_CONFIG_PATH=/opt/csw/lib/amd64/pkgconfig:/usr/lib/amd64/pkgconfig
@@ -5156,25 +5195,24 @@
   -R$@{JAVA_HOME@}/lib/amd64/server -ljvm"
 @end example
 @noindent
-With Solaris Studio 12.[23] on Sparc, @code{FCLIBS} needs to be
+With version 12.3 on Sparc, @code{FCLIBS} needs to be
 
 @example
-FCLIBS="-lfai -lfai2 -lfsu"
+FCLIBS="-lfai -lfui -lfai2 -lfsu"
 @end example
 
 @noindent
 (and possibly other Fortran libraries, but this suffices for the
 packages currently on CRAN).
 
-Currently @cputype{amd64} and @cputype{sparcv9} builds work
-out-of-the-box with Sun Studio 12u1 but not Solaris Studio 12.2 and
-12.3: @file{libRblas.so} and @file{lapack.so} are generated with code
-that causes relocation errors (which is being linked in from the Fortran
-libraries).  This means that building 64-bit @R{} as a shared library
-may be impossible with Solaris Studio >= 12.2.  For a standard build the
+Builds of @cputype{amd64} and @cputype{sparcv9} stopped working at
+Solaris Studio 12.3: @file{libRblas.so} and @file{lapack.so} are
+generated with code that causes relocation errors (code which is being
+linked in from the Fortran libraries).  This means that building 64-bit
+@R{} as a shared library may be impossible.  For a standard build the
 trick seems to be to manually set @code{FLIBS} to avoid the troublesome
 libraries.  For example, on @cputype{amd64} set in @file{config.site}
-something like
+something like (for 12.3)
 
 @example
 FLIBS_IN_SO="-R/opt/solarisstudio12.3/lib/amd64
@@ -5190,7 +5228,7 @@
  /opt/solarisstudio12.3/prod/lib/sparc/64/libfsu.so.1"
 @end example
 
-By default the Solaris Studio compilers do not by default conform to the C99
+By default the Oracle compilers do not by default conform to the C99
 standard (appendix F 8.9) on the return values of functions such as
 @code{log}: use @option{-xlibmieee} to ensure this.
 
@@ -5198,7 +5236,7 @@
 performance: @option{-xtarget=native} (in @code{CFLAGS} etc) tunes the
 compilation to the current machine.
 
-Using @code{-xlibmil} in @code{CFLAGS} and @code{-xlibmil} in
+Using @code{-xlibmil} in @code{CFLAGS} or @code{-xlibmil} in
 @code{FFLAGS} allows more system mathematical functions to be inlined.
 
 
@@ -5216,7 +5254,7 @@
 packages (notably @CRANpkg{mgcv} on @cputype{x86}) failed to compile at
 higher optimization levels with version 12.3.
 
-The Solaris Studio compilers provide several implementations of the
+The Oracle compilers provide several implementations of the
 C++98 standard which select both the set of headers and a C++ runtime
 library.  These are selected by the @option{-library} flag, which as it
 is needed for both compiling and linking is best specified as part of
@@ -5229,14 +5267,6 @@
 needs to be.  Occasionally the option @option{-library=stlport4,Crun}
 has been needed.
 
-Several @acronym{CRAN} packages using C++ need the more liberal
-interpretation given by adding
-
-@example
-CXXFLAGS="-features=tmplrefstatic"
-@end example
-
-
 @cindex BLAS library
 @cindex LAPACK library
 
@@ -5274,67 +5304,78 @@
 
 The notes here are for @command{gcc} set up to use the Solaris linker:
 it can also be set up to use GNU @command{ld}, but that has not been
-tested.
+tested.  The tests were for compilers from the OpenCSW repository:
+Solaris systems often come with much older compilers installed under
+@file{/usr/sfw/bin} and in the past an explicit path such as
+@command{/opt/csw/gcc4/bin/gcc} was need to select the latest version of the
+compilers (which at the time of writing were linked from
+@file{/opt/csw/bin} to @file{/opt/csw/gcc4/bin}).  One of @option{-m32}
+or @option{-m64} will be the default and could be omitted, but it is not
+easy to find out which.
 
-Compilation for a 32-bit Sparc target with @command{gcc}@tie{}4.9.2
-needed
+Compilation for an @cputype{x86} target with @command{gcc}@tie{}4.9.2 or
+@command{gcc}@tie{}5.2.0 needed
 
 @example
-CPPFLAGS=-I/opt/csw/include
-LDFLAGS="-L/opt/csw/gcc4/lib -L/opt/csw/lib"
+CC="gcc -m32"
+CPPFLAGS="-I/opt/csw/include -I/usr/local/include"
+F77="gfortran -m32"
+CXX="g++ -m32"
+FC="gfortran -m32"
+LDFLAGS="-L/opt/csw/lib -L/usr/local/lib"
 @end example
 
-@noindent
-and for a 64-bit Sparc target
+For an @cputype{amd64} target we used 
+
 @example
 CC="gcc -m64"
+CPPFLAGS="-I/opt/csw/include -I/usr/local/include"
 F77="gfortran -m64"
+FPICFLAGS=-fPIC
 CXX="g++ -m64"
 FC=$F77
-CPPFLAGS=-I/opt/csw/include
-LDFLAGS="-L/opt/csw/gcc4/lib/sparcv9 -L/opt/csw/lib/sparcv9"
+FCPICFLAGS=$FPICFLAGS
+LDFLAGS="-L/opt/csw/lib/amd64 -L/usr/local/lib/amd64"
 @end example
 
-@noindent
-Note that paths such as @file{/opt/csw/gcc4/lib/sparcv9} may need to
-be in the
+Note that paths such as @file{/opt/csw/lib}, @file{/usr/local/lib/amd64}
+and @file{/opt/csw/lib/amd64} may need to be in the
 @enindex LD_LIBRARY_PATH
 @env{LD_LIBRARY_PATH} during configuration.
 
-The compilation can be tuned to a particular cpu: the @acronym{CRAN}
-check system uses @code{-mtune=niagara2}.
 
-Compilation for an @cputype{x86} target with @command{gcc}@tie{}4.9.2
-needed
+Compilation for a 32-bit Sparc target with @command{gcc}@tie{}4.9.2 or
+@command{gcc}@tie{}5.2.0 needed
 
 @example
-CC="/opt/csw/gcc4/bin/gcc -m32"
+CC="gcc -m32"
+F77="gfortran -m32"
+CXX="g++ -m32"
+FC=$F77
 CPPFLAGS="-I/opt/csw/include -I/usr/local/include"
-F77="/opt/csw/gcc4/bin/gfortran -m32"
-CXX="/opt/csw/gcc4/bin/g++ -m32"
-FC="/opt/csw/gcc4/bin/gfortran -m32"
-LDFLAGS="-L/opt/csw/gcc4/lib -L/opt/csw/lib -L/usr/local/lib"
+LDFLAGS="-L/opt/csw/lib -L/usr/local/lib"
 @end example
 
 @noindent
-(@code{-L/opt/csw/lib} is needed since TeX Live was built using
-32-bit @code{gcc}, and we need @file{/opt/csw/lib} in
-@code{R_LD_LIBRARY_PATH}.)
-
-For an @cputype{amd64} target with @command{gcc}@tie{}4.9.2
-we used 
-
+and for a 64-bit Sparc target
 @example
-CC="/opt/csw/gcc4/bin/gcc -m64"
-CPPFLAGS="-I/opt/csw/include -I/usr/local/include"
-F77="/opt/csw/gcc4/bin/gfortran -m64"
-FPICFLAGS=-fPIC
-CXX="/opt/csw/gcc4/bin/g++ -m64"
+CC="gcc -m64"
+F77="gfortran -m64"
+CXX="g++ -m64"
 FC=$F77
-FCPICFLAGS=$FPICFLAGS
-LDFLAGS="-L/opt/csw/gcc4/lib/amd64 -L/opt/csw/lib/amd64"
+CPPFLAGS="-I/opt/csw/include -I/usr/local/include"
+LDFLAGS="-L/opt/csw/lib/sparcv9 -L/usr/local/lib/sparcv9"
 @end example
 
+Note that paths such as @file{/opt/csw/lib}, @file{/usr/local/lib/sparcv9} or
+@file{/opt/csw/lib/sparcv9} may need to be in the
+@enindex LD_LIBRARY_PATH
+@env{LD_LIBRARY_PATH} during configuration.
+
+The compilation can be tuned to a particular cpu: the @acronym{CRAN}
+check system uses @code{-mtune=niagara2}.
+
+
 @node AIX, FreeBSD, Solaris, Platform notes
 @section AIX
 @cindex AIX
@@ -5489,8 +5530,7 @@
 There have been few recent reports on FreeBSD.
 
 There is a `port' at @uref{https://www.freebsd.org/ports/math.html}, for
-@R{} 3.0.2 at the time of writing.  Davor Cubranic reported some success
-on x86_64 FreeBSD 10.2 for @R{} 3.2.2.
+@R{} 3.3.1 at the time of writing.
 
 Use of ICU for collation and the @command{configure} option
 @option{--with-internal-tzcode} are desirable workarounds.
@@ -5502,7 +5542,9 @@
 Ingo Feinerer installed @R{} version 3.2.2 on OpenBSD 5.8 arch
 @cputype{amd64} (their name for @cputype{x86_64}).  Details of the build
 (and patches applied) are at
-@uref{http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/math/R/}.
+@uref{http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/math/R/}. (Downgrading
+the @code{zlib} requirement to 1.2.3 is against the advice of the @R{}
+developers.)
 
 @node Cygwin, New platforms, OpenBSD, Platform notes
 @section Cygwin
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/doc/manual/R-exts.texi R-patched/doc/manual/R-exts.texi
--- R-release-orig/doc/manual/R-exts.texi	2016-04-18 00:15:04.000000000 +0200
+++ R-patched/doc/manual/R-exts.texi	2016-07-28 17:45:10.262242023 +0200
@@ -2040,13 +2040,11 @@
 Include the appropriate macro in @code{PKG_CFLAGS}, @code{PKG_CPPFLAGS}
 and so on, and also in @code{PKG_LIBS}.  C/C++ code that needs to be
 conditioned on the use of OpenMP can be used inside @code{#ifdef
-_OPENMP}: note that some toolchains used for @R{} (including many of
-those using @command{clang}@footnote{Some builds of @command{clang}
-3.7.0 have support for OpenMP 3.1 (this is built by default as from
-3.8.0), but even if the compiler has, the @code{libomp} library may not
-be installed.  At the time of writing Apple builds of @command{clang}
-for OS X had no support.})  have no OpenMP support at all, not even
-@file{omp.h}.
+_OPENMP}: note that some toolchains used for @R{} (including that of
+OS X and some others using @command{clang}@footnote{Default builds of
+@command{clang} 3.8.0 and later have support for OpenMP, but the
+@code{libomp} library may not be installed.}) have no OpenMP support at
+all, not even @file{omp.h}.
 
 For example, a package with C code written for OpenMP should have in
 @file{src/Makevars} the lines
@@ -2056,9 +2054,10 @@
 PKG_LIBS = $(SHLIB_OPENMP_CFLAGS)
 @end example
 
-Note that the macro @code{SHLIB_OPENMP_CXXFLAGS} applies to the C++98
-compiler and not necessarily to the C++11 compiler: users of the latter
-should do their own @command{configure} checks.
+Note that the macro @code{SHLIB_OPENMP_CXXFLAGS} applies to the default
+C++ compiler and not necessarily to the C++11 compiler: users of the
+latter should do their own @command{configure} checks (an example is
+available in CRAN package @CRANpkg{ARTP2}).
 
 Some care is needed when compilers are from different families which may
 use different OpenMP runtimes (e.g.@: @command{clang} @emph{vs} GCC
@@ -2075,16 +2074,24 @@
 as the C compiler will be used to link the package code (and there is no
 guarantee that this will work everywhere).
 
-
-There is nothing to say what version of OpenMP is supported: version 3.0
-(May 2008) is supported by recent versions of the Linux, Windows and
-Solaris platforms, but portable packages cannot assume that end users
-have recent versions.  OS X currently uses Apple builds of
-@command{clang} with no OpenMP support.
-
-The performance of OpenMP varies substantially between platforms.  Both
-the Windows and earlier Apple OS X implementations have substantial
-overheads and are only beneficial if quite substantial tasks are run in
+@c http://openmp.org/wp/openmp-compilers/
+@c clang 3.8.x reports 201307 but has full suport only for 3.1 (201111)
+There is nothing@footnote{In most implementations the
+@code{_OPENMP} macro has value a date which can be mapped to an OpenMP
+version: for example, value @code{201307} is the date of version 4.0
+(July 2013). However this may be used to denote the latest version which
+is partially supported, not that which is fully implemented.}  to say
+what version of OpenMP is supported: version 3.1 is supported by recent
+versions of the Linux, Windows and Solaris platforms, but portable
+packages cannot assume that end users have recent versions.  macOS
+currently uses Apple builds of @command{clang} with no OpenMP support
+(even if invoked as @command{gcc} and despite the @command{man} page
+including the flag @option{-fopenmp} for that command).
+
+The performance of OpenMP varies substantially between platforms.  The
+Windows implementation has substantial overheads@footnote{as did the
+GCC-based Apple implementation, but not if the Intel OpenMP runtime was
+used.}, so is only beneficial if quite substantial tasks are run in
 parallel.  Also, on Windows new threads are started with the
 default@footnote{Windows default, not MinGW-w64 default.} FPU control
 word, so computations done on OpenMP threads will not make use of
@@ -2110,7 +2117,7 @@
 overall number of threads available to OpenMP in the @R{} process: this
 can be done via environment variable @env{OMP_THREAD_LIMIT}, where
 implemented.@footnote{Which it was at the time of writing with GCC,
-Solaris Studio, Intel and Clang 3.7.x compilers.}  Alternatively, the
+Solaris Studio, Intel and Clang compilers.}  Alternatively, the
 number of threads per region can be limited by the environment variable
 @env{OMP_NUM_THREADS} or API call @code{omp_set_num_threads}, or,
 better, for the regions in your code as part of their
@@ -2215,7 +2222,7 @@
 
 $(SHLIB): Csdp/lib/libsdp.a
 
-Csdp/lib/libsdp.a
+Csdp/lib/libsdp.a:      
 	@@(cd Csdp/lib && $(MAKE) libsdp.a \
 	  CC="$(CC)" CFLAGS="$(CFLAGS) $(CPICFLAGS)" AR="$(AR)" RANLIB="$(RANLIB)")
 @end smallexample
@@ -2225,7 +2232,10 @@
 -m64 -std=gnu99"}.  Several authors have forgotten about parallel makes:
 the static library in the subdirectory must be made before the shared
 object (@code{$(SHLIB)}) and so the latter must depend on the former.
-Others forget the need for position-independent code.
+Others forget the need@footnote{A few OSes (AIX, IRIX, Windows) do not
+need special flags for such code, but most do---although compilers will
+often generate PIC code when not asked to do so.} for
+position-independent code.
 
 We really do not recommend using @file{src/Makefile} instead of
 @file{src/Makevars}, and as the example above shows, it is not
@@ -2436,13 +2446,13 @@
 C++11 (often known as C++14) was approved in August 2014.
 
 What standard a C++ compiler aims to support can be hard to determine.
-As from version 6 (to be released in 2016), @command{g++} will default
-to C++14: earlier versions aim to support C++03 with many extensions
+As from version 6, @command{g++} defaults to C++14 (with GNU
+extensions): earlier versions aim to support C++03 with many extensions
 (including support for TR1).  @command{clang} with its
 native@footnote{Some distributions, notably Debian, have supplied a
 build of @command{clang} with @command{g++}'s headers and library.
 Conversely, Apple's command named @command{g++} is based on
-@command{clang} using @code{libcxx}.}  @code{libcxx} headers and library
+@command{clang} using @code{libc++}.}  @code{libc++} headers and library
 includes many C++11 features, and does not support TR1.
 
 Since version 3.1.0, @R{} has provided support for C++11 in packages in
@@ -2494,25 +2504,32 @@
 @xref{Customizing package compilation, , Customizing package compilation,
 R-admin, R Installation and Administration}.
 @end ifset
-If C++11 support is not available then these macros are both
-empty. Otherwise, @samp{CXX1X} defaults to the same value as the C++
-compiler @samp{CXX} and the flag @samp{CXX1XSTD} defaults to
-@option{-std=c++11} or @option{-std=c++0x} (the latter on Windows).  It
-is possible to specify @samp{CXX1X} to be a distinct compiler just for
-C++11--using packages, e.g.@: @command{g++} on Solaris.  Note however
-that different C++ compilers (and even different versions of the same
-compiler) often differ in their ABI so their outputs can rarely be
-mixed. By setting @samp{CXX1XSTD} it is also possible to choose a
-different dialect of the standard, such as @option{-std=gnu++11}, or
-enable support for the 2014 revision using something like
-@option{-std=c++14} or @option{-std=c++1y}.
-
-As noted above, support for C++11 varies across platforms.  The default
-compiler on Windows is GCC 4.6.x and supports the @option{-std=c++0x}
-flag and some C++11 features (see
-@uref{https://gcc.gnu.org/@/gcc-4.6/@/cxx0x_status.html}).  On some
+If C++11 support is not available then these macros are both empty; if
+it is available by default, @samp{CXX1X} defaults to @samp{CXX} and
+@samp{CXX1XSTD} is empty . Otherwise, @samp{CXX1X} defaults to the same
+value as the C++ compiler @samp{CXX} and the flag @samp{CXX1XSTD}
+defaults to @option{-std=c++11} or @option{-std=c++0x} (the latter on
+Windows).  It is possible to specify @samp{CXX1X} to be a distinct
+compiler just for C++11--using packages, e.g.@: @command{g++} on
+Solaris.  Note however that different C++ compilers (and even different
+versions of the same compiler) often differ in their ABI so their
+outputs can rarely be mixed. By setting @samp{CXX1XSTD} it is also
+possible to choose a different dialect of the standard such as
+@option{-std=gnu++11}, or enable support for the 2014 revision using
+something like @option{-std=c++14} or @option{-std=c++1y}.
+
+As noted above, support for C++11 varies across platforms: on some
 platforms, it may be possible or necessary to select a different
 compiler for C++11, @emph{via} personal or site @file{Makevars} files.
+Note too that support for C++11 may include C++14 support, so that
+including emulation of C++14 features should be done by something like
+@example
+#if __cplusplus >= 201402L
+using std::make_unique;
+#else
+// your emulation
+#endif
+@end example
 
 There is no guarantee that C++11 can be used in a package in combination
 with any other compiled language (even C), as the C++11 compiler may be
@@ -2525,7 +2542,7 @@
 @example
 CXX1X=`"$@{R_HOME@}/bin/R" CMD config CXX1X`
 CXX1XSTD=`"$@{R_HOME@}/bin/R" CMD config CXX1XSTD`
-CXX="$(CXX1X) $(CXX1XSTD)"
+CXX="$@{CXX1X@} $@{CXX1XSTD@}"
 CXXFLAGS=`"$@{R_HOME@}/bin/R" CMD config CXX1XFLAGS`
 AC_LANG(C++)
 @end example
@@ -2533,6 +2550,14 @@
 @noindent
 (paying attention to all the quotes required).
 
+If you want to compile C++11 code in a subdirectory, make sure you pass
+down the macros to specify that compiler, e.g.@: in @file{src/Makevars}
+@example
+sublibs:
+         @@(cd libs && $(MAKE) \
+            CXX="$(CXX1X) $(CXX1XSTD)" CXXFLAGS="$(CXX1XFLAGS) $(CXX1XPICFLAGS)")
+@end example
+
 @node Checking and building packages, Writing package vignettes, Configure and cleanup, Creating R packages
 @section Checking and building packages
 
@@ -4028,13 +4053,13 @@
 @uref{http://pubs.opengroup.org/@/onlinepubs/@/9699919799/@/utilities/@/make.html};
 anything not documented there should be regarded as an extension to be
 avoided.)  Commonly misused GNU extensions are conditional inclusions
-(@code{ifeq} and the like), @code{$@{shell ...@}} and @code{$@{wildcard
-...@}}, and the use of @code{+=}@footnote{This was apparently introduced
-in SunOS 4, and is available elsewhere @emph{provided} it is surrounded
-by spaces.}  and @code{:=}.  Also, the use of @code{$<} other than in
-implicit rules is a GNU extension, as is the @code{$^} macro.
-Unfortunately makefiles which use GNU extensions often run on other
-platforms but do not have the intended results.
+(@code{ifeq} and the like), @code{$@{shell ...@}}, @code{$@{wildcard
+...@}} and similar, and the use of @code{+=}@footnote{This was
+apparently introduced in SunOS 4, and is available elsewhere
+@emph{provided} it is surrounded by spaces.} and @code{:=}.  Also, the
+use of @code{$<} other than in implicit rules is a GNU extension, as is
+the @code{$^} macro.  Unfortunately makefiles which use GNU extensions
+often run on other platforms but do not have the intended results.
 
 The use of @code{$@{shell ...@}} can be avoided by using backticks, e.g.@:
 
@@ -4106,15 +4131,19 @@
 R-admin, R Installation and Administration},
 @end ifset
 
-Although there is a 2011 version of the C++ standard, even partial
-implementations are not universally available.  Portable C++ code needs
-to follow the 1998 standard (and not use features from C99).  See also
-@ref{Using C++11 code} to specify a C++11 compiler.
+Although there is a 2011 version of the C++ standard, implementations
+are not universally available.  Portable C++ code needs to follow the
+1998 standard (and not use features from C99).  See also @ref{Using
+C++11 code} to specify a C++11 compiler where available.
 
 If you use FORTRAN 77, @code{ftnchek}
 (@uref{http://www.dsm.fordham.edu/@/~ftnchek/}) provides thorough testing
 of conformance to the standard.
 
+If using Fortran 95 with the GNU compiler, use the flag
+@option{-std=f95} which rejects most GNU extensions and features from
+later standards.
+
 Not all common @R{} platforms conform to the expected standards, e.g.@:
 C99 for C code.  One common area of problems is the @code{*printf}
 functions where Windows does not support @code{%lld}, @code{%Lf} and
@@ -4123,6 +4152,20 @@
 integers can usually be converted to doubles for output.
 
 @item
+@command{R CMD check} performs some checks for non-portable
+compiler/linker flags in @file{src/Makevars}.  However, it cannot check
+the meaning of such flags, and some are commonly accepted but with
+compiler-specific meanings.  There are other non-portable flags such as
+@option{-isystem} which are not checked and @file{src/Makefile} files
+and makefiles is sub-directories are also not checked.  As a comment in
+the code says
+
+@quotation
+It is hard to think of anything apart from @option{-I*} and @option{-D*}
+that is safe for general use @dots{}
+@end quotation
+
+@item
 Do be very careful with passing arguments between @R{}, C and
 @acronym{FORTRAN} code.  In particular, @code{long} in C will be 32-bit
 on some @R{} platforms (including 64-bit Windows), but 64-bit on most
@@ -4281,6 +4324,15 @@
 However, some Linux distributions micro-package, so locales defined by
 @pkg{glibc} (including these examples) may not be installed.
 
+@item
+Avoid spaces in file names, not least as they can cause difficulties for
+external tools.  A recent example was a package with a @pkg{knitr}
+vignette that used spaces in plot names: this caused some versions of
+@command{pandoc} to fail with a baffling error message.
+@c msmtools in June 2016 failed with pandoc 1.12 but not 1.16.
+
+Non-ASCII filenames can also cause problems (particularly in non-UTF-8
+locales).
 @end itemize
 
 Do be careful in what your tests (and examples) actually test.  Bad
@@ -4572,17 +4624,26 @@
 these functions for integer arguments.
 
 @item
-The GNU compilers support a large number of non-portable extensions.  For
-example, @code{INFINITY} (which is in C99 but not C++98), for which @R{}
-provides the portable @code{R_PosInf} (and @code{R_NegInf} for
-@code{-INFINITY}).  And @code{NAN} is just one NaN value: in @R{} code
-@code{NA_REAL} is usually what is intended, but @code{R_NaN} is also
-available.
+The GNU C/C++ compilers support a large number of non-portable
+extensions.  For example, @code{INFINITY} (which is in C99 but not
+C++98), for which @R{} provides the portable @code{R_PosInf} (and
+@code{R_NegInf} for @code{-INFINITY}).  And @code{NAN} is just one NaN
+value: in @R{} code @code{NA_REAL} is usually what is intended, but
+@code{R_NaN} is also available.
 
 Some (but not all) extensions are listed at
 @uref{https://gcc.gnu.org/@/onlinedocs/@/gcc/@/C-Extensions.html} and
 @uref{https://gcc.gnu.org/@/onlinedocs/@/gcc/@/C_002b_002b-Extensions.html}.
 
+The GNU Fortran compiler also supports a large number of non-portable
+extensions, the most commonly encountered one being
+@code{ISNAN}@footnote{There is a portable way to do this in Fortran 2003
+(@code{ieee_is_nan()} in module @code{ieee_arithmetic}), but ironically
+that is not supported in the commonly-used versions 4.x of GNU
+Fortran.  A pretty robust alternative is to test @code{if(my_var /=
+my_var)}.}.  Some are listed at
+@uref{https://gcc.gnu.org/@/onlinedocs/@/gfortran/@/Extensions-implemented-in-GNU-Fortran.html}
+
 @item
 Including C-style headers in C++ code is not portable.  Including the
 legacy header@footnote{which often is the same as the header included by
@@ -4648,6 +4709,23 @@
 @item
 @code{typedef}s in OS headers can conflict with those in the package: an
 example is @code{index_t} defined in @file{<sys/types.h>} on Solaris.
+
+@item
+If you use OpenMP, check carefully that you have followed the advice in
+the subsection on @ref{OpenMP support}.  In particular, any use of
+OpenMP in C/C++ code will need to use
+@example
+#ifdef _OPENMP
+# include <omp.h>
+#endif
+@end example
+@noindent
+Any use of OpenMP functions, e.g.@: @code{omp_set_num_threads} also
+needs to be conditioned.
+
+And do not hardocde @option{-lgomp}: not only is that specific to the
+GCC family of compilers, using the correct linker flag sets up the
+run-time path to the library.
 @end itemize
 
 Some additional information for C++ is available at
@@ -8011,11 +8089,12 @@
 @command{clang} see
 @uref{http://clang.llvm.org/@/docs/@/UsersManual.html#controlling-code-generation}.@footnote{or
 the user manual for your version of @command{clang}, e.g.@:
-@uref{http://llvm.org/@/releases/@/3.6.2@//tools/@/docs/@/UsersManual.html}.}
+@uref{http://llvm.org/@/releases/@/3.8.1/@/tools/@/docs/@/UsersManual.html} or
+@uref{http://llvm.org/@/releases/@/3.8.0/@/tools/@/clang/@/docs/@/UsersManual.html}.}
 The current set for @command{clang} is (on a single line):
 @example
 -fsanitize=alignment,bool,bounds,enum,float-cast-overflow,
-float-divide-by-zero,function,integer-divide-by-zero,non-null-attribute,
+float-divide-by-zero,function,integer-divide-by-zero,nonnull-attribute,
 null,object-size,return,returns-nonnull-attribute,shift,
 signed-integer-overflow,unreachable,vla-bound,vptr
 @end example
@@ -8066,7 +8145,7 @@
 GCC: 5.3 supports
 @example
 -fsanitize=alignment,bool,bounds,enum,float-cast-overflow,
-integer-divide-by-zero,non-null-attribute,null,object-size,
+integer-divide-by-zero,nonnull-attribute,null,object-size,
 return,returns-nonnull-attribute,shift,signed-integer-overflow,
 unreachable,vla-bound,vptr
 @end example
@@ -8081,7 +8160,7 @@
 and @code{vptr} checks produced many warnings on GCC's own C++ headers,
 so should be disabled.
 
-GCC 6 will add
+GCC 6 added
 @example
 -fsanitize=bounds-strict
 @end example
@@ -9338,7 +9417,7 @@
 The @R{} for Windows @acronym{FAQ} (@file{rw-FAQ}) contains details of how
 to compile this example under Windows.
 
-Earlier version of this example used C++ iostreams: this is best
+Earlier versions of this example used C++ iostreams: this is best
 avoided.  There is no guarantee that the output will appear in the @R{}
 console, and indeed it will not on the @R{} for Windows console.  Use
 @R{} code or the C entry points (@pxref{Printing}) for all I/O if at all
@@ -9351,24 +9430,44 @@
 they include system headers@footnote{Even including C system headers in
 such a block has caused compilation errors.}).  It may not be possible
 to include some @R{} headers as they in turn include system header files
-that may cause conflicts---if this happens, try defining
-@samp{NO_C_HEADERS} before including the @R{} headers, and include C++
-versions (such as @samp{cmath} and @samp{cstdlib}) of the appropriate
-headers yourself before the @R{} headers. (Headers @file{R.h} and
-@file{Rmath.h} support @samp{NO_C_HEADERS}: the legacy header @file{S.h}
-does not.  Header @file{Rinternals.h} does as from @R{} 3.3.0.)
-
-By default header @file{Rmath.h} includes @file{math.h} or @file{cmath}.
-Header @file{R.h} includes
+that may cause conflicts---if this happens, first try including the
+appropriate headers yourself before the @R{} headers, then try defining
+@samp{NO_C_HEADERS} before including the @R{} headers.  (Headers
+@file{R.h}, @file{Rinternals.h} and @file{Rmath.h} support
+@samp{NO_C_HEADERS}: @file{cstddef} needs to be included before
+@file{R.h} if @samp{NO_C_HEADERS} is defined.)
+
+Legacy header @file{S.h} cannot be used with C++.
+
+@subsection External C++ code
+
+Quite a lot of external C++ software is header-only (e.g.@: most of the
+Boost `libraries' including all those supplied by package @CRANpkg{BH},
+and most of Armadillo as supplied by package @CRANpkg{RcppArmadillo})
+and so is compiled when an @R{} package which uses it is installed.
+This causes few problems.
+
+A small number of external libraries used in @R{} packages have a C++
+interface to a library of compiled code, e.g.@: packages @CRANpkg{rgdal}
+and @CRANpkg{rjags}.  This raises many more problems!  The C++ interface
+uses name-mangling which depends on the compiler, version and even C++
+defines, so requires the package C++ code to be compiled in exactly the
+same way as the library (and what that was is often undocumented).
+Examples include use of @command{g++} @emph{vs} @command{clang++} or
+Solaris' @command{CC}, and the two ABIs available for C++11 in
+@command{g++} with different defaults for GCC 4.9 and 5.x in some Linux
+distributions.
 
-@example
-limits.h math.h stddef.h stdio.h stdlib.h string.h
-@end example
+Even fewer external libraries use C++ internally but present a C
+interface, such as @CRANpkg{rgeos}.  These require the C++ runtime
+library to be linked into the package's shared object/DLL, and this is
+best done by including a dummy C++ file in the package sources.
+
+There is a recent trend to link to the C++ interfaces offered by C
+software such as @pkg{hdf5}, @pkg{pcre} and @pkg{ImageMagick}.  Their C
+interfaces are much preferred for portability (and can be used from C++
+code).
 
-@noindent
-or their C++ equivalents directly or indirectly, and either
-@file{stddef.h} or @file{cstddef} needs to be included before @file{R.h}
-if @samp{NO_C_HEADERS} is defined.
 
 @node Fortran I/O, Linking to other packages, Interfacing C++ code, System and foreign language interfaces
 @section Fortran I/O
@@ -9381,12 +9480,12 @@
 In the past most Fortran compilers implemented I/O on top of the C I/O
 system and so the two interworked successfully.  This was true of
 @command{g77}, but it is less true of @command{gfortran} as used in
-@code{gcc 4.y.z}.  In particular, any package that makes use of Fortran
-I/O will when compiled on Windows interfere with C I/O: when the Fortran
-I/O is initialized (typically when the package is loaded) the C
+@command{gcc} 4 and later.  In particular, any package that makes use of
+Fortran I/O will when compiled on Windows interfere with C I/O: when the
+Fortran I/O is initialized (typically when the package is loaded) the C
 @code{stdout} and @code{stderr} are switched to LF line endings.
-(Function @code{init} in file @file{src/modules/lapack/init_win.c} shows how to
-mitigate this.)
+(Function @code{init} in file @file{src/modules/lapack/init_win.c} shows
+how to mitigate this.)
 
 @node Linking to other packages, Handling R objects in C, Fortran I/O, System and foreign language interfaces
 @section Linking to other packages
@@ -12853,10 +12952,12 @@
 @samp{HAVE_VISIBILITY_ATTRIBUTE} is defined -- this is checked when @R{}
 is configured and recorded in header @file{Rconfig.h} (included by
 @file{R_ext/Visibility.h}).  It is generally defined on modern
-Unix-alikes with a recent compiler, but not supported on OS X nor
-Windows.  Minimizing the visibility of symbols in a shared library will
-both speed up its loading (unlikely to be significant) and reduce the
-possibility of linking to other entry points of the same name.
+Unix-alikes with a recent compiler@footnote{It is defined by the Intel
+compilers, but also hides unsatisfied references and so cannot be used
+with @R{}.}, but not supported on OS X nor Windows.  Minimizing the
+visibility of symbols in a shared library will both speed up its loading
+(unlikely to be significant) and reduce the possibility of linking to
+other entry points of the same name.
 
 C/C++ entry points prefixed by @code{attribute_hidden} will not be
 visible in the shared object.  There is no comparable mechanism for
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/m4/R.m4 R-patched/m4/R.m4
--- R-release-orig/m4/R.m4	2016-03-17 00:02:06.000000000 +0100
+++ R-patched/m4/R.m4	2016-07-26 17:45:07.911186000 +0200
@@ -1,6 +1,6 @@
 ### R.m4 -- extra macros for configuring R		-*- Autoconf -*-
 ###
-### Copyright (C) 1998-2015 R Core Team
+### Copyright (C) 1998-2016 R Core Team
 ###
 ### This file is part of R.
 ###
@@ -3261,7 +3261,7 @@
 fi
 AC_MSG_CHECKING([whether bzip2 support suffices])
 if test "x${have_bzlib}" = xyes; then
-  AC_MSG_RESULT([no])
+  AC_MSG_RESULT([yes])
   LIBS="-lbz2 ${LIBS}"
 else
   AC_MSG_ERROR([bzip2 library and headers are required])
@@ -3665,13 +3665,19 @@
 ])# R_CHECK_FUNCS
 
 ## R_GCC4_VISIBILITY
-## Sets up suitable macros for visibility attributes in gcc4/gfortran
+## Sets up suitable macros for visibility attributes in gcc/gfortran
+## Also accepted on clang (which defines __GNUC__). 
+## Intel also defines __GNUC__ but is excluded below, and
+## Solaris <= 12.4 rejected -Werror, but 12.5 did not.
 AC_DEFUN([R_GCC4_VISIBILITY],
 [AC_CACHE_CHECK([whether __attribute__((visibility())) is supported],
                 [r_cv_visibility_attribute],
 [cat > conftest.c <<EOF
 int foo __attribute__ ((visibility ("hidden"))) = 1;
 int bar __attribute__ ((visibility ("default"))) = 1;
+#ifndef __GNUC__
+# error unsupported compiler
+#endif
 EOF
 r_cv_visibility_attribute=no
 if AC_TRY_COMMAND(${CC-cc} -Werror -S conftest.c -o conftest.s 1>&AS_MESSAGE_LOG_FD); then
@@ -3698,8 +3704,8 @@
     C_VISIBILITY="-fvisibility=hidden"
   fi
 fi
-## Need to exclude Intel compilers, where this does not work.
-## The flag is documented, and is effective but also hides
+## Need to exclude Intel compilers, where this does not work correctly.
+## The flag is documented and is effective, but also hides
 ## unsatisfied references. We cannot test for GCC, as icc passes that test.
 case  "${CC}" in
   ## Intel compiler: note that -c99 may have been appended
@@ -4165,7 +4171,7 @@
     CURL_LIBS=`${CURL_CONFIG} --libs`
   fi
 fi
-r_save_CPPFLAGS="${CPPLAGS}"
+r_save_CPPFLAGS="${CPPFLAGS}"
 CPPFLAGS="${CURL_CPPFLAGS} ${CPPFLAGS}"
 r_save_LIBS="${LIBS}"
 LIBS="${CURL_LIBS} ${LIBS}"
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/m4/openmp.m4 R-patched/m4/openmp.m4
--- R-release-orig/m4/openmp.m4	2015-09-09 00:15:06.000000000 +0200
+++ R-patched/m4/openmp.m4	2016-07-12 02:41:42.349421000 +0200
@@ -1,6 +1,7 @@
 # This file is part of Autoconf.			-*- Autoconf -*-
 # Programming languages support.
 # Copyright (C) 2001-2012 Free Software Foundation, Inc.
+# Copyright (C) 2015-2016 R Core Team
 
 # This file is part of Autoconf.  This program is free
 # software; you can redistribute it and/or modify it under the
@@ -27,7 +28,7 @@
 # Franc,ois Pinard, Karl Berry, Richard Pixley, Ian Lance Taylor,
 # Roland McGrath, Noah Friedman, david d zuhn, and many others.
 
-# [a small part, modified for clang and icc in 2015.]
+# [a small part, modified for clang and Intel in 2015,6.]
 
 
 # R_OPENMP
@@ -53,18 +54,21 @@
 	 [ac_cv_prog_[]_AC_LANG_ABBREV[]_openmp='none needed'],
 	 [ac_cv_prog_[]_AC_LANG_ABBREV[]_openmp='unsupported'
 	  dnl Try these flags:
-	  dnl   GCC >= 4.2           -fopenmp
-	  dnl   clang 3.7.0	     -fopenmp=libomp
+	  dnl   GCC >= 4.2, clang 3.8 -fopenmp
+	  dnl   clang 3.7.x	      -fopenmp=libomp
 	  dnl   (-fopenmp is accepted but does not work)
-	  dnl   SolarisStudio C      -xopenmp
-	  dnl   Intel C              -openmp (deprecated)
-	  dnl   icc (ca 2015) 	     -qopenmp
-	  dnl   SGI C, PGI C         -mp
-	  dnl   Tru64 Compaq C       -omp
-	  dnl   IBM C (AIX, Linux)   -qsmp=omp
-          dnl   Cray CCE             -homp
-          dnl   NEC SX               -Popenmp
-          dnl   Lahey Fortran (Linux)  --openmp
+	  dnl   Oracle C, Fortran     -xopenmp
+          dnl   (also accepts -fopenmp as from 12.4)
+	  dnl   Intel C, Fortran      -qopenmp
+	  dnl   Intel                 -openmp (deprecated)
+          dnl   (https://software.intel.com/en-us/node/581863,
+	  dnl    https://software.intel.com/en-us/node/525020)
+	  dnl   SGI C, PGI C          -mp
+	  dnl   Tru64 Compaq C        -omp
+	  dnl   IBM C (AIX, Linux)    -qsmp=omp
+          dnl   Cray CCE              -homp
+          dnl   NEC SX                -Popenmp
+          dnl   Lahey Fortran (Linux) --openmp
 	  dnl If in this loop a compiler is passed an option that it doesn't
 	  dnl understand or that it misinterprets, the AC_LINK_IFELSE test
 	  dnl will fail (since we know that it failed without the option),
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/share/make/check_vars_out.mk R-patched/share/make/check_vars_out.mk
--- R-release-orig/share/make/check_vars_out.mk	2013-07-19 00:05:06.000000000 +0200
+++ R-patched/share/make/check_vars_out.mk	2016-06-23 17:45:10.000000000 +0200
@@ -5,7 +5,7 @@
 ##                sprintf("\t@echo %sFLAGS: \"$(%sFLAGS)\"",
 ##                        prefixes, prefixes)))
 
-all:
+makevars_test:
 	@echo PKG_CPPFLAGS: "$(PKG_CPPFLAGS)"
 	@echo PKG_CFLAGS: "$(PKG_CFLAGS)"
 	@echo PKG_CXXFLAGS: "$(PKG_CXXFLAGS)"
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/share/texmf/bibtex/bib/RJournal.bib R-patched/share/texmf/bibtex/bib/RJournal.bib
--- R-release-orig/share/texmf/bibtex/bib/RJournal.bib	2016-06-21 00:15:14.000000000 +0200
+++ R-patched/share/texmf/bibtex/bib/RJournal.bib	2016-07-24 02:41:25.000000000 +0200
@@ -316,6 +316,45 @@
 
 
 
+@Article{brown:,
+  author       = {Patrick Brown}, 
+  title        = {Maps, coordinate reference systems and visualising geographic data with mapmisc (was: Simple and flexible maps of geospatial data with mapmisc)}, 
+  journal      = {The R Journal},
+  year         = ,
+  volume       = ,
+  number       = ,
+  pages        = {},  
+  month        = ,
+  url          = {http://journal.r-project.org/archive/-/brown.pdf}
+
+
+
+@Article{winslow-chamberlain-appling-etal:,
+  author       = {Luke A Winslow and Scott Chamberlain and Alison P Appling and Jordan S Read}, 
+  title        = {sbtools: A package connecting R to cloud-based data for collaborative online research}, 
+  journal      = {The R Journal},
+  year         = ,
+  volume       = ,
+  number       = ,
+  pages        = {},  
+  month        = ,
+  url          = {http://journal.r-project.org/archive/-/winslow-chamberlain-appling-etal.pdf}
+
+
+
+@Article{wais:,
+  author       = {Kamil Wais}, 
+  title        = {Gender Prediction Methods Based on First Names with genderizeR}, 
+  journal      = {The R Journal},
+  year         = ,
+  volume       = ,
+  number       = ,
+  pages        = {},  
+  month        = ,
+  url          = {http://journal.r-project.org/archive/-/wais.pdf}
+
+
+
 @Article{RJournal:2009-1:Chambers,
   author       = {John M. Chambers}, 
   title        = {Facets of {R}}, 
@@ -2683,7 +2722,7 @@
 
 
 @Article{cattaneo-calonico-titiunik:2015,
-  author       = {Matias D. Cattaneo and Sebastian Calonico and Rocio Titiunik}, 
+  author       = {Sebastian Calonico and Matias D. Cattaneo and Rocio Titiunik}, 
   title        = {{rdrobust}: An {R} Package for Robust Inference in Regression-Discontinuity Designs}, 
   journal      = {The R Journal},
   year         = 2015,
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/extra/tre/tre-internal.h R-patched/src/extra/tre/tre-internal.h
--- R-release-orig/src/extra/tre/tre-internal.h	2016-03-17 00:02:11.000000000 +0100
+++ R-patched/src/extra/tre/tre-internal.h	2016-07-27 17:45:09.153857000 +0200
@@ -54,6 +54,10 @@
   Solaris-sparcv9   WCHAR_MAX = INT32_MAX
   Linux amd64       WCHAR_MAX = INT32_MAX
 */
+/*
+   [U]INT32_MAX need to be declared: this is a C99 header which we assume
+ */
+#include <stdint.h>
 #if WCHAR_MAX == UINT32_MAX
 # define TRE_CHAR_MAX INT32_MAX
 #else
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/gnuwin32/COPYRIGHTS.win R-patched/src/gnuwin32/COPYRIGHTS.win
--- R-release-orig/src/gnuwin32/COPYRIGHTS.win	2016-06-10 00:15:06.000000000 +0200
+++ R-patched/src/gnuwin32/COPYRIGHTS.win	2016-06-21 17:45:10.000000000 +0200
@@ -140,6 +140,6 @@
 
 A binary build may contain compiled components from ICU, whose license
 can be viewed at
-http://source.icu-project.org/repos/icu/icu/trunk/license.html: the
+http://source.icu-project.org/repos/icu/icu/trunk/LICENSE: the
 sources and binary libraries used are available at
 http://www.stats.ox.ac.uk/pub/Rtools/ .
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/gnuwin32/getline/getline.c R-patched/src/gnuwin32/getline/getline.c
--- R-release-orig/src/gnuwin32/getline/getline.c	2013-03-06 00:02:42.000000000 +0100
+++ R-patched/src/gnuwin32/getline/getline.c	2016-07-07 17:45:10.000000000 +0200
@@ -147,9 +147,9 @@
       st = r.Event.KeyEvent.dwControlKeyState;
       vk = r.Event.KeyEvent.wVirtualKeyCode;
       if (r.Event.KeyEvent.bKeyDown) {
-	if (vk == VK_MENU && (st & LEFT_ALT_PRESSED)) { /* VK_MENU is
+        AltIsDown = (st & LEFT_ALT_PRESSED);
+	if (vk == VK_MENU && AltIsDown) { /* VK_MENU is
 							   Alt or AltGr */
-	  AltIsDown = 1;
 	  nAlt = 0;
 	  bbb  = 0;
 	} 
Binary files R-release-orig/src/library/Recommended/KernSmooth.tgz and R-patched/src/library/Recommended/KernSmooth.tgz differ
Binary files R-release-orig/src/library/Recommended/KernSmooth_2.23-15.tar.gz and R-patched/src/library/Recommended/KernSmooth_2.23-15.tar.gz differ
Binary files R-release-orig/src/library/Recommended/MASS.tgz and R-patched/src/library/Recommended/MASS.tgz differ
Binary files R-release-orig/src/library/Recommended/MASS_7.3-45.tar.gz and R-patched/src/library/Recommended/MASS_7.3-45.tar.gz differ
Binary files R-release-orig/src/library/Recommended/boot.tgz and R-patched/src/library/Recommended/boot.tgz differ
Binary files R-release-orig/src/library/Recommended/boot_1.3-18.tar.gz and R-patched/src/library/Recommended/boot_1.3-18.tar.gz differ
Binary files R-release-orig/src/library/Recommended/class.tgz and R-patched/src/library/Recommended/class.tgz differ
Binary files R-release-orig/src/library/Recommended/class_7.3-14.tar.gz and R-patched/src/library/Recommended/class_7.3-14.tar.gz differ
Binary files R-release-orig/src/library/Recommended/nnet.tgz and R-patched/src/library/Recommended/nnet.tgz differ
Binary files R-release-orig/src/library/Recommended/nnet_7.3-12.tar.gz and R-patched/src/library/Recommended/nnet_7.3-12.tar.gz differ
Binary files R-release-orig/src/library/Recommended/rpart.tgz and R-patched/src/library/Recommended/rpart.tgz differ
Binary files R-release-orig/src/library/Recommended/rpart_4.1-10.tar.gz and R-patched/src/library/Recommended/rpart_4.1-10.tar.gz differ
Binary files R-release-orig/src/library/Recommended/spatial.tgz and R-patched/src/library/Recommended/spatial.tgz differ
Binary files R-release-orig/src/library/Recommended/spatial_7.3-11.tar.gz and R-patched/src/library/Recommended/spatial_7.3-11.tar.gz differ
Binary files R-release-orig/src/library/Recommended/survival.tgz and R-patched/src/library/Recommended/survival.tgz differ
Only in R-release-orig/src/library/Recommended: survival_2.39-4.tar.gz
Binary files R-release-orig/src/library/Recommended/survival_2.39-5.tar.gz and R-patched/src/library/Recommended/survival_2.39-5.tar.gz differ
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/R/match.R R-patched/src/library/base/R/match.R
--- R-release-orig/src/library/base/R/match.R	2016-03-17 00:03:52.000000000 +0100
+++ R-patched/src/library/base/R/match.R	2016-06-29 02:41:11.000000000 +0200
@@ -34,6 +34,8 @@
     .Internal(pmatch(as.character(x), as.character(table), nomatch,
                      duplicates.ok))
 
+# "utils::hasName(x, name)" is defined to be the same as "name %in% names(x)",
+# so change it if this changes.
 `%in%`  <- function(x, table) match(x, table, nomatch = 0L) > 0L
 
 match.arg <- function (arg, choices, several.ok = FALSE)
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/R/merge.R R-patched/src/library/base/R/merge.R
--- R-release-orig/src/library/base/R/merge.R	2015-08-26 00:17:26.000000000 +0200
+++ R-patched/src/library/base/R/merge.R	2016-07-24 02:41:26.000000000 +0200
@@ -1,7 +1,7 @@
 #  File src/library/base/R/merge.R
 #  Part of the R package, https://www.R-project.org
 #
-#  Copyright (C) 1995-2012 The R Core Team
+#  Copyright (C) 1995-2016 The R Core Team
 #
 #  This program is free software; you can redistribute it and/or modify
 #  it under the terms of the GNU General Public License as published by
@@ -169,9 +169,11 @@
         res <- cbind(x, y)
 
         if (sort)
-            res <- res[if(all.x || all.y) ## does NOT work
-                       do.call("order", x[, seq_len(l.b), drop = FALSE])
-            else sort.list(bx[m$xi]),, drop = FALSE]
+	    res <- res[if(all.x || all.y) {
+			   x <- x[, seq_len(l.b), drop = FALSE]
+			   attributes(x) <- NULL
+			   do.call("order", x)
+		       } else sort.list(bx[m$xi]),, drop = FALSE]
     }
     attr(res, "row.names") <- .set_row_names(nrow(res))
     res
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/R/namespace.R R-patched/src/library/base/R/namespace.R
--- R-release-orig/src/library/base/R/namespace.R	2016-05-12 00:15:06.000000000 +0200
+++ R-patched/src/library/base/R/namespace.R	2016-07-19 02:41:08.000000000 +0200
@@ -296,7 +296,7 @@
                            lapply(type,
                                   function(sym) {
                                       varName <- paste0(fixes[1L], sym$name, fixes[2L])
-                                      if(exists(varName, envir = env))
+                                      if(exists(varName, envir = env, inherits = FALSE))
                                           warning(gettextf("failed to assign RegisteredNativeSymbol for %s to %s since %s is already defined in the %s namespace",
                                                            sym$name, varName, varName, sQuote(package)),
                                                   domain = NA, call. = FALSE)
@@ -319,7 +319,7 @@
                        ## maintain the original names.
                        varName <- names(symNames)[i]
                        origVarName <- symNames[i]
-                       if(exists(varName, envir = env))
+                       if(exists(varName, envir = env, inherits = FALSE))
                            if(origVarName != varName)
                                warning(gettextf("failed to assign NativeSymbolInfo for %s to %s since %s is already defined in the %s namespace",
                                                 origVarName, varName, varName, sQuote(package)),
@@ -438,7 +438,10 @@
         codename <- strsplit(package, "_", fixed = TRUE)[[1L]][1L]
         codeFile <- file.path(pkgpath, "R", codename)
         if (file.exists(codeFile)) {
+	    # The code file has been converted to the native encoding
+	    save.enc <- options(encoding = "native.enc")
             res <- try(sys.source(codeFile, env, keep.source = keep.source))
+	    options(save.enc)
             if(inherits(res, "try-error"))
                 stop(gettextf("unable to load R code in package %s",
                               sQuote(package)), call. = FALSE, domain = NA)
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/R/sample.R R-patched/src/library/base/R/sample.R
--- R-release-orig/src/library/base/R/sample.R	2015-08-26 00:17:22.000000000 +0200
+++ R-patched/src/library/base/R/sample.R	2016-07-19 17:45:10.000000000 +0200
@@ -18,7 +18,7 @@
 
 sample <- function(x, size, replace = FALSE, prob = NULL)
 {
-    if(length(x) == 1L && is.numeric(x) && x >= 1) {
+    if(length(x) == 1L && is.numeric(x) && is.finite(x) && x >= 1) {
 	if(missing(size)) size <- x
 	sample.int(x, size, replace, prob)
     } else {
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/Extract.Rd R-patched/src/library/base/man/Extract.Rd
--- R-release-orig/src/library/base/man/Extract.Rd	2015-08-26 00:18:13.000000000 +0200
+++ R-patched/src/library/base/man/Extract.Rd	2016-06-26 02:41:08.000000000 +0200
@@ -1,6 +1,6 @@
 % File src/library/base/man/Extract.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 1995-2014 R Core Team
+% Copyright 1995-2016 R Core Team
 % Copyright 2002-2012 The R Foundation
 % Distributed under GPL 2 or later
 
@@ -36,7 +36,7 @@
 x[i] <- value
 x[i, j, \dots] <- value
 x[[i]] <- value
-x$i <- value
+x$name <- value
 }
 
 \arguments{
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/Extremes.Rd R-patched/src/library/base/man/Extremes.Rd
--- R-release-orig/src/library/base/man/Extremes.Rd	2015-08-26 00:17:53.000000000 +0200
+++ R-patched/src/library/base/man/Extremes.Rd	2016-07-28 17:45:09.000000000 +0200
@@ -1,16 +1,16 @@
 % File src/library/base/man/Extremes.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 1995-2013 R Core Team
+% Copyright 1995-2016 R Core Team
 % Distributed under GPL 2 or later
 
 \name{Extremes}
+\title{Maxima and Minima}
 \alias{max}
 \alias{min}
 \alias{pmax}
 \alias{pmin}
 \alias{pmax.int}
 \alias{pmin.int}
-\title{Maxima and Minima}
 \description{
   Returns the (parallel) maxima and minima of the input values.
 }
@@ -56,7 +56,8 @@
   elements of all the arguments and so on.  Shorter inputs (of non-zero
   length) are recycled if necessary.  Attributes (see
   \code{\link{attributes}}: such as \code{\link{names}} or
-  \code{\link{dim}}) are copied from the first argument (if applicable).
+  \code{\link{dim}}) are copied from the first argument (if applicable,
+  e.g., \emph{not} for an \code{S4} object).
 
   \code{pmax.int} and \code{pmin.int} are faster internal versions only
   used when all arguments are atomic vectors and there are no classes:
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/c.Rd R-patched/src/library/base/man/c.Rd
--- R-release-orig/src/library/base/man/c.Rd	2015-08-26 00:17:41.000000000 +0200
+++ R-patched/src/library/base/man/c.Rd	2016-07-28 17:45:09.000000000 +0200
@@ -1,6 +1,6 @@
 % File src/library/base/man/c.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 1995-2013 R Core Team
+% Copyright 1995-2016 R Core Team
 % Distributed under GPL 2 or later
 
 \name{c}
@@ -29,6 +29,12 @@
   components (such names and calls) are treated as one-element lists
   which cannot be unlisted even if \code{recursive = TRUE}.
 
+  %% Should we say "(currently)" ?
+  Note that \code{\link{factor}}s are treated only via their
+  internal \code{\link{integer}} codes; one proposal has been to use
+\preformatted{   c.factor <- function(..., recursive=TRUE) unlist(list(...), recursive=recursive)}
+  if factor concatenation by \code{c()} should give a \code{\link{factor}}.
+
   \code{c} is sometimes used for its side effect of removing attributes
   except names, for example to turn an array into a vector.
   \code{as.vector} is a more intuitive way to do this, but also drops
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/exists.Rd R-patched/src/library/base/man/exists.Rd
--- R-release-orig/src/library/base/man/exists.Rd	2015-08-26 00:17:59.000000000 +0200
+++ R-patched/src/library/base/man/exists.Rd	2016-06-29 02:41:11.000000000 +0200
@@ -1,6 +1,6 @@
 % File src/library/base/man/exists.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 1995-2015 R Core Team
+% Copyright 1995-2016 R Core Team
 % Distributed under GPL 2 or later
 
 \name{exists}
@@ -87,7 +87,8 @@
   Wadsworth & Brooks/Cole.
 }
 \seealso{
-  \code{\link{get}}.  For quite a different kind of \dQuote{existence}
+  \code{\link{get}} and \code{\link{hasName}}.  For quite a different 
+  kind of \dQuote{existence}
   checking, namely if function arguments were specified,
   \code{\link{missing}};
   and for yet a different kind, namely if a file exists,
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/extSoftVersion.Rd R-patched/src/library/base/man/extSoftVersion.Rd
--- R-release-orig/src/library/base/man/extSoftVersion.Rd	2015-08-26 00:18:11.000000000 +0200
+++ R-patched/src/library/base/man/extSoftVersion.Rd	2016-06-21 17:45:10.000000000 +0200
@@ -1,6 +1,6 @@
 % File src/library/base/man/extSoftVersion.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 2014-2015 R Core Team
+% Copyright 2014-2016 R Core Team
 % Distributed under GPL 2 or later
 
 \name{extSoftVersion}
@@ -32,6 +32,8 @@
   \item{TRE}{The version of \code{libtre} in use.}
   \item{iconv}{The implementation and version of the \code{iconv}
   library in use (if known).}
+\item{readline}{The version of \code{readline} in use (if any,
+  otherwise \code{""}).}
 
   Note that the values for \code{bzlib} and \code{pcre} normally contain
   a date as well as the version number, and that for \code{tre} includes
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/files.Rd R-patched/src/library/base/man/files.Rd
--- R-release-orig/src/library/base/man/files.Rd	2016-05-21 00:15:05.000000000 +0200
+++ R-patched/src/library/base/man/files.Rd	2016-07-28 17:45:09.000000000 +0200
@@ -211,7 +211,7 @@
 
   \code{\link{Sys.glob}} to expand wildcards in file specifications.
 
-  \code{\link{file_test}}, \code{\link{Sys.readlink}}.
+  \code{\link{file_test}}, \code{\link{Sys.readlink}} (for \sQuote{symlink}s).
 
   \url{https://en.wikipedia.org/wiki/Hard_link} and
   \url{https://en.wikipedia.org/wiki/Symbolic_link} for the concepts of
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/srcfile.Rd R-patched/src/library/base/man/srcfile.Rd
--- R-release-orig/src/library/base/man/srcfile.Rd	2015-08-26 00:18:13.000000000 +0200
+++ R-patched/src/library/base/man/srcfile.Rd	2016-07-28 17:45:09.000000000 +0200
@@ -1,9 +1,10 @@
 % File src/library/base/man/srcfile.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 1995-2014 R Core Team
+% Copyright 1995-2016 R Core Team
 % Distributed under GPL 2 or later
 
 \name{srcfile}
+\title{References to Source Files and Code}
 \alias{srcfile}
 \alias{srcfilecopy}
 \alias{getSrcLines}
@@ -25,9 +26,12 @@
 \alias{open.srcfilealias}
 \alias{close.srcfilealias}
 
-\title{References to source files}
 \description{
-These functions are for working with source files.
+  These functions are for working with source files and more generally
+  with \dQuote{source references} (\code{"srcref"}), i.e., references to
+  source code.  The resulting data is used for printing and source level
+  debugging, and is typically available in interactive \R sessions,
+  namely when \code{\link{options}(keep.source = TRUE)}.
 }
 \usage{
 srcfile(filename, encoding = getOption("encoding"), Enc = "unknown")
@@ -115,7 +119,7 @@
   method will read the associated source file to obtain the text
   corresponding to the reference.  If the \code{to} argument is given,
   it should be a second \code{srcref} that follows the first, in the
-  same file; they will be treated as one reference to the whole 
+  same file; they will be treated as one reference to the whole
   range.  The exact behaviour depends on the
   class of the source file.  If the source file inherits from
   class \code{srcfilecopy}, the lines are taken from the saved copy
@@ -150,7 +154,9 @@
 \code{srcref} returns a \code{srcref} object.
 }
 \seealso{
-\code{\link{getSrcFilename}} for extracting information from a source reference.
+  \code{\link{getSrcFilename}} for extracting information from a source
+  reference, or \code{\link{removeSource}} to remove it from a
+  (non-primitive) function (aka \sQuote{closure}).
 }
 \author{Duncan Murdoch}
 \examples{\donttest{ # has timestamp
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/tapply.Rd R-patched/src/library/base/man/tapply.Rd
--- R-release-orig/src/library/base/man/tapply.Rd	2016-03-17 00:04:00.000000000 +0100
+++ R-patched/src/library/base/man/tapply.Rd	2016-07-28 17:45:09.000000000 +0200
@@ -1,6 +1,6 @@
 % File src/library/base/man/tapply.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 1995-2015 R Core Team
+% Copyright 1995-2016 R Core Team
 % Distributed under GPL 2 or later
 
 \name{tapply}
@@ -16,16 +16,18 @@
 }
 \arguments{
   \item{X}{an atomic object, typically a vector.}
-  \item{INDEX}{list of one or more factors, each of same length as
-    \code{X}.  The elements are coerced to factors by \code{\link{as.factor}}.}
+  \item{INDEX}{a \code{\link{list}} of one or more \code{\link{factor}}s,
+    each of same length as \code{X}.  The elements are coerced to
+    factors by \code{\link{as.factor}}.}
   \item{FUN}{the function to be applied, or \code{NULL}.  In the case of
     functions like \code{+}, \code{\%*\%}, etc., the function name must
     be backquoted or quoted.  If \code{FUN} is \code{NULL}, tapply
     returns a vector which can be used to subscript the multi-way array
     \code{tapply} normally produces.}
   \item{\dots}{optional arguments to \code{FUN}: the Note section.}
-  \item{simplify}{If \code{FALSE}, \code{tapply} always returns an array
-    of mode \code{"list"}.  If \code{TRUE} (the default), then if
+  \item{simplify}{logical; if \code{FALSE}, \code{tapply} always returns
+    an array of mode \code{"list"}; in other words, a \code{\link{list}}
+    with a \code{\link{dim}} attribute.  If \code{TRUE} (the default), then if
     \code{FUN} always returns a scalar, \code{tapply} returns an array
     with the mode of the scalar.}
 }
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/base/man/timezones.Rd R-patched/src/library/base/man/timezones.Rd
--- R-release-orig/src/library/base/man/timezones.Rd	2016-02-21 00:15:05.000000000 +0100
+++ R-patched/src/library/base/man/timezones.Rd	2016-07-11 17:45:10.000000000 +0200
@@ -164,7 +164,7 @@
   mainland France) and switched to \samp{GMT+0} in 1911.
 
   %% it is a ksh script so could well pop up elsewhere.
-  Some systems (notably Linux) have \command{tzselect} command which
+  Some systems (notably Linux) have a \command{tzselect} command which
   allows the interactive selection of a supported time zone name.
 }
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grDevices/R/prettyDate.R R-patched/src/library/grDevices/R/prettyDate.R
--- R-release-orig/src/library/grDevices/R/prettyDate.R	2016-05-27 00:15:05.000000000 +0200
+++ R-patched/src/library/grDevices/R/prettyDate.R	2016-07-13 12:28:48.000000000 +0200
@@ -221,6 +221,7 @@
     } else if(md < 15) { ## e.g., southern hemisphere, seen 14
         at2$mday <- md + 14L # consistent w (1 -> 15) in 1st case; ok even in Feb.
     }
+    at2$isdst <- -1L
     at2 <- rbind(at, as.POSIXct(at2), deparse.level = 0L)
     structure(at2[i], class = class(at), tzone = attr(at, "tzone"))
 }
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grDevices/man/pretty.Date.Rd R-patched/src/library/grDevices/man/pretty.Date.Rd
--- R-release-orig/src/library/grDevices/man/pretty.Date.Rd	2016-04-24 00:15:05.000000000 +0200
+++ R-patched/src/library/grDevices/man/pretty.Date.Rd	2016-07-13 12:28:48.000000000 +0200
@@ -28,6 +28,9 @@
   length one, when \code{min.n = 0} and there is only one unique
   \code{x}.
 }
+%% \note{The current algorithm is geared towards small \code{n}, not
+%% larger than say, 50 or 100.
+%% }
 \value{
   A vector (of the suitable class) of locations, with attribute
   \code{"labels"} giving corresponding formatted character labels.
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grDevices/src/colors.c R-patched/src/library/grDevices/src/colors.c
--- R-release-orig/src/library/grDevices/src/colors.c	2015-08-26 00:18:22.000000000 +0200
+++ R-patched/src/library/grDevices/src/colors.c	2016-07-17 17:45:08.000000000 +0200
@@ -109,6 +109,8 @@
     double f, p, q, t;
     int i;
 
+    if (!R_FINITE(h) || !R_FINITE(s) || !R_FINITE(v)) 
+	error(_("inputs must be finite")); 
     f = modf(h * 6.0, &t);
     i = ((int) t) % 6;
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grDevices/src/devices.c R-patched/src/library/grDevices/src/devices.c
--- R-release-orig/src/library/grDevices/src/devices.c	2015-08-26 00:18:22.000000000 +0200
+++ R-patched/src/library/grDevices/src/devices.c	2016-07-17 17:45:08.000000000 +0200
@@ -85,20 +85,25 @@
 SEXP devnext(SEXP args)
 {
     checkArity_length;
-    return ScalarInteger( nextDevice(INTEGER(CAR(args))[0] - 1) + 1 );
+    int nxt = INTEGER(CAR(args))[0];
+    if (nxt == NA_INTEGER) error(_("NA argument is invalid"));
+    return ScalarInteger( nextDevice(nxt - 1) + 1 );
 }
 
 SEXP devprev(SEXP args)
 {
     checkArity_length;
-    return ScalarInteger( prevDevice(INTEGER(CAR(args))[0] - 1) + 1 );
+    int prev = INTEGER(CAR(args))[0];
+    if (prev == NA_INTEGER) error(_("NA argument is invalid"));
+    return ScalarInteger( prevDevice(prev - 1) + 1 );
 }
 
 SEXP devset(SEXP args)
 {
     checkArity_length;
-    int devNum = INTEGER(CAR(args))[0] - 1;
-    return ScalarInteger( selectDevice(devNum) + 1 );
+    int devNum = INTEGER(CAR(args))[0];
+    if (devNum == NA_INTEGER) error(_("NA argument is invalid"));
+    return ScalarInteger( selectDevice(devNum - 1) + 1 );
 }
 
 SEXP devoff(SEXP args)
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grDevices/tests/ps-tests.Rout.save R-patched/src/library/grDevices/tests/ps-tests.Rout.save
--- R-release-orig/src/library/grDevices/tests/ps-tests.Rout.save	2016-06-15 00:15:05.000000000 +0200
+++ R-patched/src/library/grDevices/tests/ps-tests.Rout.save	2016-07-21 17:45:10.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -95,4 +95,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.393   0.022   0.402 
+  0.336   0.017   0.343 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grDevices/tests/xfig-tests.Rout.save R-patched/src/library/grDevices/tests/xfig-tests.Rout.save
--- R-release-orig/src/library/grDevices/tests/xfig-tests.Rout.save	2016-06-15 00:15:05.000000000 +0200
+++ R-patched/src/library/grDevices/tests/xfig-tests.Rout.save	2016-07-21 17:45:10.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -102,4 +102,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.159   0.024   0.172 
+  0.168   0.021   0.179 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grid/R/unit.R R-patched/src/library/grid/R/unit.R
--- R-release-orig/src/library/grid/R/unit.R	2016-03-17 00:02:44.000000000 +0100
+++ R-patched/src/library/grid/R/unit.R	2016-07-21 17:45:10.000000000 +0200
@@ -200,7 +200,7 @@
 }
 
 valid.units <- function(units) {
-  .Call(validUnits, units)
+  .Call(L_validUnits, units)
 }
 
 as.character.unit <- function(x, ...) {
Binary files R-release-orig/src/library/grid/inst/doc/displaylist.pdf and R-patched/src/library/grid/inst/doc/displaylist.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/frame.pdf and R-patched/src/library/grid/inst/doc/frame.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/grid.pdf and R-patched/src/library/grid/inst/doc/grid.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/grobs.pdf and R-patched/src/library/grid/inst/doc/grobs.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/interactive.pdf and R-patched/src/library/grid/inst/doc/interactive.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/locndimn.pdf and R-patched/src/library/grid/inst/doc/locndimn.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/moveline.pdf and R-patched/src/library/grid/inst/doc/moveline.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/nonfinite.pdf and R-patched/src/library/grid/inst/doc/nonfinite.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/plotexample.pdf and R-patched/src/library/grid/inst/doc/plotexample.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/rotated.pdf and R-patched/src/library/grid/inst/doc/rotated.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/saveload.pdf and R-patched/src/library/grid/inst/doc/saveload.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/sharing.pdf and R-patched/src/library/grid/inst/doc/sharing.pdf differ
Binary files R-release-orig/src/library/grid/inst/doc/viewports.pdf and R-patched/src/library/grid/inst/doc/viewports.pdf differ
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grid/src/register.c R-patched/src/library/grid/src/register.c
--- R-release-orig/src/library/grid/src/register.c	2015-08-26 00:15:37.000000000 +0200
+++ R-patched/src/library/grid/src/register.c	2016-07-21 17:45:10.000000000 +0200
@@ -72,7 +72,7 @@
     {"L_locator", (DL_FUNC) &L_locator, 0},
     {"L_convert", (DL_FUNC) &L_convert, 4},
     {"L_layoutRegion", (DL_FUNC) &L_layoutRegion, 2},
-    {"validUnits", (DL_FUNC) &validUnits, 1},
+    {"L_validUnits", (DL_FUNC) &validUnits, 1},
     {"L_getGPar", (DL_FUNC) &L_getGPar, 0},
     {"L_setGPar", (DL_FUNC) &L_setGPar, 1},
     {"L_circleBounds", (DL_FUNC) &L_circleBounds, 4},
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/grid/tests/testls.Rout.save R-patched/src/library/grid/tests/testls.Rout.save
--- R-release-orig/src/library/grid/tests/testls.Rout.save	2016-06-15 00:15:05.000000000 +0200
+++ R-patched/src/library/grid/tests/testls.Rout.save	2016-07-21 17:45:10.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -340,4 +340,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.213   0.023   0.225 
+  0.222   0.018   0.228 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/R/ClassExtensions.R R-patched/src/library/methods/R/ClassExtensions.R
--- R-release-orig/src/library/methods/R/ClassExtensions.R	2016-03-17 00:03:40.000000000 +0100
+++ R-patched/src/library/methods/R/ClassExtensions.R	2016-07-16 17:45:08.000000000 +0200
@@ -383,7 +383,7 @@
         if(exti@distance == 1)
             next # asserted that this was done by makeExtends
         what <- superClasses[[i]]
-        whatDef <- getClassDef(what, where)
+        whatDef <- getClassDef(what, package=packageSlot(exti))
         if(is.null(whatDef) # but shouldn't happen,
            || !isXS3Class(whatDef))
             next
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/R/RClassUtils.R R-patched/src/library/methods/R/RClassUtils.R
--- R-release-orig/src/library/methods/R/RClassUtils.R	2016-03-17 00:03:40.000000000 +0100
+++ R-patched/src/library/methods/R/RClassUtils.R	2016-07-16 17:45:08.000000000 +0200
@@ -31,7 +31,7 @@
             return(TRUE)
         ## does the class extend a known non-virtual class?
         for(what in en) {
-            enDef <- getClassDef(what, where)
+            enDef <- getClassDef(what, package=packageSlot(extends[[what]]))
             if(!is.null(enDef) && identical(enDef@virtual, FALSE))
                 return(FALSE)
         }
@@ -84,8 +84,8 @@
         if(identical(what, "VIRTUAL")) {
             ## the class is virtual, and the prototype usually NULL
 ##            virtual <- TRUE
-        } else if(isClass(what, where = where)) {
-            cli <- getClass(what, where = where)
+        } else if(isClass(what, where=packageSlot(exti))) {
+            cli <- getClassDef(what, package=packageSlot(exti))
             slotsi <- names(cli@slots)
             pri <- cli@prototype
             ## once in a while
@@ -170,7 +170,7 @@
                          paste(extra, collapse=", ")),
                 domain = NA)
     ## now check the elements of the prototype against the class definition
-    slotDefs <- getSlots(ClassDef); slotNames <- names(slotDefs)
+    slotDefs <- ClassDef@slots; slotNames <- names(slotDefs)
     pnames <- names(attributes(prototype))
     pnames <- pnames[!is.na(match(pnames, slotNames))]
     check <- rep.int(FALSE, length(pnames))
@@ -458,8 +458,8 @@
     addCond <- function(xpr, prev)
         if(length(prev)) substitute(P && N, list(P = prev, N = xpr)) else xpr
     C <- if(dropVirtual) {
-        ## NB the default 'where' in getClass() may depend on specific superClass:
-        isVirtualExt <- function(x) getClass(x@superClass)@virtual
+             isVirtualExt <- function(x)
+                 getClassDef(x@superClass, package=packageSlot(x))@virtual
         quote(!isVirtualExt(exti))
     } else expression()
     if(directOnly) C <- addCond(quote(length(exti@by) == 0), C)
@@ -477,8 +477,11 @@
         Class@contains
     else if(isClass(Class, where = where))
         getClass(Class, where = where)@contains
-    supcl <- .selectSuperClasses(ext) ## maybe  simpleOnly = FALSE or use as argument?
-    unique(unlist(lapply(lapply(supcl, getClassDef), slotNames), use.names=FALSE))
+    supcl <- .selectSuperClasses(ext, namesOnly=FALSE) ## maybe  simpleOnly = FALSE or use as argument?
+    supdefs <- lapply(supcl, function(s) {
+        getClassDef(s@superClass, package=packageSlot(s))
+    })
+    unique(unlist(lapply(supdefs, slotNames), use.names=FALSE))
     ## or just the non-simplified part (*with* names):
     ##     lapply(sapply(supcl, getClassDef, simplify=FALSE), slotNames)
 }
@@ -1073,7 +1076,7 @@
         for(i in seq_along(contains)) {
             obji <- contains[[i]]
             cli <- contains[[i]]@superClass
-            cliDef <- getClassDef(cli, where)
+            cliDef <- getClassDef(cli, package=packageSlot(obji))
             ## don't override existing relations
             ## TODO:  have a metric that picks the "closest" relationship
             if(!extends(classDef2, cliDef))
@@ -1098,8 +1101,8 @@
     what <- names(ext)
     for(i in seq_along(ext)) { # note that this loops only over the original ext
         by <- what[[i]]
-        if(isClass(by, where = where)) {
-            byDef <- getClass(by, where = where)
+        if(isClass(by, where = packageSlot(ext[[i]]))) {
+            byDef <- getClassDef(by, package=packageSlot(ext[[i]]))
             exti <-  slot(byDef, slotName)
             coni <- attr(exti, "conflicts") # .resolveSuperclasses makes this
             if(superClassCase && length(coni) > 0) {
@@ -1164,8 +1167,9 @@
     for(i in seq_along(ext)) {
         by <- what[[i]]
         ## report only the direct superclass from which inconsistencies are inherited
-        if(identical(ext[[i]]@distance, 1) && isClass(by, where = where)) {
-            byDef <- getClass(by, where = where)
+        wherei <- packageSlot(ext[[i]])
+        if(identical(ext[[i]]@distance, 1) && isClass(by, where = wherei)) {
+            byDef <- getClassDef(by, package=wherei)
             exti <-  byDef@contains
             coni <- attr(exti, "conflicts") # .resolveSuperclasses makes this
             if( length(coni) > 0) {
@@ -2071,7 +2075,7 @@
     subNames <- names(subs)
     for(i in seq_along(subs)) {
         what <- subNames[[i]]
-        subDef <- getClassDef(what, env)
+        subDef <- getClassDef(what, package=packageSlot(subs[[i]]))
         if(is.null(subDef))
             warning(gettextf("undefined subclass %s of class %s; definition not updated",
                              .dQ(what), .dQ(def@className)))
@@ -2133,7 +2137,7 @@
     if(length(classDef@contains)) {
         superclasses <- names(classDef@contains)
         for(what in superclasses) {
-            cdef <- .getClassFromCache(what, resolve.confl = "all")
+            cdef <- .getClassFromCache(what, classWhere, resolve.confl = "all")
 	    if(is(cdef, "classRepresentation"))
 		.removeSubClass(what, Class, cdef)
 	    else if(is.list(cdef))
@@ -2203,8 +2207,8 @@
 
 ## remove superclass from  definition of class in the cache & in environments
 ## on search list
-.removeSuperClass <- function(class, superclass, resolve.msg = TRUE) {
-    cdef <- .getClassFromCache(class, where, resolve.msg=resolve.msg)
+.removeSuperClass <- function(class, superclass) {
+    cdef <- getClassDef(class)
     if(is.null(cdef)) {}
     else {
         newdef <- .deleteSuperClass(cdef, superclass)
@@ -2220,7 +2224,7 @@
     mname <- classMetaName(class)
     for(where in evv) {
         if(!is.null(cdef <- where[[mname]])) {
-            newdef <- .deleteSuperClass(cdef, superclass, resolve.msg=resolve.msg)
+            newdef <- .deleteSuperClass(cdef, superclass)
             if(!is.null(newdef)) {
               assignClassDef(class, newdef,  where, TRUE)
               ## message("deleted ",superclass, " from ",class, "in environment")
@@ -2230,13 +2234,13 @@
     NULL
 }
 
-.deleteSuperClass <- function(cdef, superclass, resolve.msg = TRUE) {
+.deleteSuperClass <- function(cdef, superclass) {
     superclasses <- cdef@contains
     ii <- match(superclass, names(superclasses), 0L)
     if(ii) {
 	cdef@contains <- superclasses[-ii]
 	for(subclass in names(cdef@subclasses))
-	    .removeSuperClass(subclass, superclass, resolve.msg=resolve.msg)
+	    .removeSuperClass(subclass, superclass)
 	cdef
     }
     else
@@ -2282,7 +2286,7 @@
         superclasses <- names(extensions)
         edges <- numeric()
         for(what in superclasses) {
-            whatDef <- getClassDef(what)
+            whatDef <- getClassDef(what, package=packageSlot(extensions[[what]]))
             ifrom <- match(what, nodes)
             if(is.null(whatDef) || is.na(ifrom))
               next
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/R/RMethodUtils.R R-patched/src/library/methods/R/RMethodUtils.R
--- R-release-orig/src/library/methods/R/RMethodUtils.R	2016-03-17 00:03:40.000000000 +0100
+++ R-patched/src/library/methods/R/RMethodUtils.R	2016-07-16 17:45:08.000000000 +0200
@@ -1720,7 +1720,7 @@
         ## else, an error
         classi <- classes[[i]]
         pkgi <- pkgs[[i]]
-        classDefi <- getClass(classi, where = where)
+        classDefi <- getClassDef(classi, where=if (pkgi == "") where else pkgi)
         if(checkDups && classi %in% multipleClasses()) { # hardly ever, we hope
             clDefsi <- get(classi, envir = .classTable)
             if(nzchar(pkgi) && pkgi %in% names(clDefsi))
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/R/SClasses.R R-patched/src/library/methods/R/SClasses.R
--- R-release-orig/src/library/methods/R/SClasses.R	2016-03-17 00:03:40.000000000 +0100
+++ R-patched/src/library/methods/R/SClasses.R	2016-07-16 17:45:08.000000000 +0200
@@ -69,6 +69,8 @@
         classDef@sealed <- FALSE # to allow setIs to work anyway; will be reset later
         assignClassDef(Class, classDef, where)
         badContains <- character()
+### FIXME: need to iterate over contains, not superclass to get
+### package for getClassDef()
         for(class2 in superClasses) {
             if(is(try(setIs(Class, class2, classDef = classDef, where = where)), "try-error"))
                 badContains <- c(badContains, class2)
@@ -233,10 +235,11 @@
 getClassDef <-
   ## Get the definition of the class supplied as a string.
   function(Class, where = topenv(parent.frame()), package = packageSlot(Class),
-           inherits = TRUE, resolve.msg = getOption("getClass.msg", default=TRUE))
+           inherits = TRUE)
 {
-    value <- if(inherits) #includes both the lookup and Class being already a definition
-	.getClassFromCache(Class, where, package=package, resolve.msg=resolve.msg)
+    value <- if(inherits)
+                 .getClassFromCache(Class, where, package=package,
+                                    resolve.confl="none")
     ## else NULL # want to force a search for the metadata in this case (Why?)
     if(is.null(value)) {
 	cname <-
@@ -246,7 +249,10 @@
 			  Class[[1L]] else Class)
 	## a string with a package slot strongly implies the class definition
 	## should be in that package.
-	if(identical(nzchar(package), TRUE)) {
+        if(is.character(where)) {
+            package <- where
+        }
+	if(isTRUE(nzchar(package))) {
 	    whereP <- .requirePackage(package)
 	    value <- get0(cname, whereP, inherits = inherits) # NULL if not existing
 	}
@@ -260,13 +266,10 @@
   ## Get the complete definition of the class supplied as a string,
   ## including all slots, etc. in classes that this class extends.
   function(Class, .Force = FALSE,
-	   where = .classEnv(Class, topenv(parent.frame()), FALSE),
-           resolve.msg = getOption("getClass.msg", default=TRUE))
+	   where = .classEnv(Class, topenv(parent.frame()), FALSE))
 {
-    value <- .getClassFromCache(Class, where, resolve.msg=resolve.msg) # the quick way
+    value <- getClassDef(Class, where)
     if(is.null(value)) {
-        value <- getClassDef(Class, where, resolve.msg=resolve.msg) # searches
-        if(is.null(value)) {
             if(!.Force)
                 stop(gettextf("%s is not a defined class",
                               dQuote(Class)),
@@ -274,7 +277,6 @@
             else
                 value <- makeClassRepresentation(Class, package = "base",
                                                  virtual = TRUE, where = where)
-        }
     }
     value
 }
@@ -377,8 +379,7 @@
 }
 
 
-removeClass <-  function(Class, where = topenv(parent.frame()),
-                         resolve.msg = getOption("removeClass.msg", default=TRUE)) {
+removeClass <-  function(Class, where = topenv(parent.frame())) {
     if(missing(where)) {
        classEnv <- .classEnv(Class, where, FALSE)
         classWhere <- findClass(Class, where = classEnv)
@@ -401,7 +402,7 @@
       subclasses <- names(classDef@subclasses)
       found <- vapply(subclasses, isClass, NA, where = where, USE.NAMES=TRUE)
       for(what in subclasses[found])
-          .removeSuperClass(what, Class, resolve.msg=resolve.msg)
+          .removeSuperClass(what, Class)
     }
     .removeSuperclassBackRefs(Class, classDef, classWhere)
     .uncacheClass(Class, classDef)
@@ -524,7 +525,7 @@
 	superClass <- exti@superClass
 	if(!exti@simple && !is(object, superClass))
 	    next ## skip conditional relations that don't hold for this object
-	superDef <- getClassDef(superClass, where = where)
+	superDef <- getClassDef(superClass, package = packageSlot(exti))
 	if(is.null(superDef)) {
 	    errors <- c(errors,
 			paste0("superclass \"", superClass,
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/R/oldClass.R R-patched/src/library/methods/R/oldClass.R
--- R-release-orig/src/library/methods/R/oldClass.R	2016-03-17 00:03:40.000000000 +0100
+++ R-patched/src/library/methods/R/oldClass.R	2016-07-16 17:45:08.000000000 +0200
@@ -273,7 +273,7 @@
 
 .addS3Class <- function(class, prototype, contains, where) {
     for(what in contains) {
-        whatDef <- getClassDef(what@superClass, where = where)
+        whatDef <- getClassDef(what@superClass, package=packageSlot(what))
         if(isXS3Class(whatDef))
           class <- c(class, attr(whatDef@prototype, ".S3Class"))
     }
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/man/findClass.Rd R-patched/src/library/methods/man/findClass.Rd
--- R-release-orig/src/library/methods/man/findClass.Rd	2015-08-26 00:17:15.000000000 +0200
+++ R-patched/src/library/methods/man/findClass.Rd	2016-07-13 12:28:48.000000000 +0200
@@ -21,8 +21,7 @@
 
 findClass(Class, where, unique = "")
 
-removeClass(Class, where,
-            resolve.msg = getOption("removeClass.msg", default=TRUE))
+removeClass(Class, where)
 
 resetClass(Class, classDef, where)
 
@@ -57,9 +56,6 @@
     environment only?  Defaults to \code{TRUE} if \code{where} is
     omitted, and to \code{FALSE} otherwise.
   }
-  \item{resolve.msg}{\code{\link{logical}} indicating if \R should
-    \code{\link{message}()} its decision if \code{Class} is found in
-    multiple namespaces and one is chosen.}
   \item{classDef}{ For \code{resetClass}, the optional class
     definition (but usually it's better for \code{Class} to be the
     class definition, and to omit \code{classDef}).
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/man/getClass.Rd R-patched/src/library/methods/man/getClass.Rd
--- R-release-orig/src/library/methods/man/getClass.Rd	2015-08-26 00:17:15.000000000 +0200
+++ R-patched/src/library/methods/man/getClass.Rd	2016-07-13 12:28:48.000000000 +0200
@@ -11,10 +11,8 @@
   Get the definition of a class.
 }
 \usage{
-getClass   (Class, .Force = FALSE, where,
-            resolve.msg = getOption("getClass.msg", default=TRUE))
-getClassDef(Class, where, package, inherits = TRUE,
-            resolve.msg = getOption("getClass.msg", default=TRUE))
+getClass   (Class, .Force = FALSE, where)
+getClassDef(Class, where, package, inherits = TRUE)
 }
 \arguments{
   \item{Class}{ the character-string name of the class, often with a
@@ -35,9 +33,6 @@
   \item{inherits}{logical; should the class definition be retrieved from
     any enclosing environment and also from the cache?  If \code{FALSE}
     only a definition in the environment \code{where} will be returned.}
-  \item{resolve.msg}{\code{\link{logical}} indicating if \R should
-    \code{\link{message}()} its decision if \code{Class} is found in
-    multiple namespaces and one is chosen.}
 }
 \details{
   Class definitions are stored in metadata objects in a package
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/methods/man/getMethod.Rd R-patched/src/library/methods/man/getMethod.Rd
--- R-release-orig/src/library/methods/man/getMethod.Rd	2015-08-26 00:17:15.000000000 +0200
+++ R-patched/src/library/methods/man/getMethod.Rd	2016-07-28 17:45:10.000000000 +0200
@@ -1,15 +1,15 @@
 % File src/library/methods/man/getMethod.Rd
 % Part of the R package, https://www.R-project.org
-% Copyright 1995-2012 R Core Team
+% Copyright 1995-2016 R Core Team
 % Distributed under GPL 2 or later
 
 \name{getMethod}
+\title{Get or Test for the Definition of a Method}
 \alias{getMethod}
 \alias{findMethod}
 \alias{existsMethod}
 \alias{selectMethod}
 \alias{hasMethod}
-\title{ Get or Test for the Definition of a Method }
 \description{
   Functions to look for a method corresponding to a given generic function and signature.
   The functions \code{getMethod} and \code{selectMethod} return the method; the functions
@@ -23,60 +23,57 @@
   contain a method for this function and signature.
 }
 \usage{
-getMethod(f, signature=character(), where, optional = FALSE,
-          mlist, fdef)
-
 existsMethod(f, signature = character(), where)
 
-findMethod(f, signature, where)
+  findMethod(f, signature, where)
 
-selectMethod(f, signature, optional = FALSE, useInherited =,
-             mlist = , fdef = , verbose = , doCache = )
+   getMethod(f, signature = character(), where, optional = FALSE, mlist, fdef)
 
-hasMethod(f, signature=character(), where)
+   hasMethod(f, signature = character(), where)
 
+selectMethod(f, signature, optional = FALSE, useInherited =,
+             mlist = , fdef = , verbose = , doCache = )
 }
 \arguments{
-  \item{f}{A  generic function or the character-string name of one.}
+  \item{f}{a generic function or the character-string name of one.}
   \item{signature}{the signature of classes to match to the arguments
-    of \code{f}.  See the details below.
-
-  }
-  \item{where}{The position or environment in which to look for the
-    method(s):  by default, the table of methods defined in the generic function itself is used.}
-
-  \item{optional}{If the selection in \code{selectMethod} does not find a valid method
-    an error is generated, unless this argument is \code{TRUE}.  In that
-    case, the value returned is \code{NULL} if no method matches.}
-  \item{mlist, fdef, useInherited, verbose, doCache}{Optional
-      arguments  to  \code{getMethod} and \code{selectMethod} for
-      internal use.  Avoid
-      these: some will work as expected and others will not, and none of them is required for normal use of the functions.
-    }
-
+    of \code{f}.  See the details below.}
+  \item{where}{the position or environment in which to look for the
+    method(s):  by default, the table of methods defined in the generic
+    function itself is used.}
+
+  \item{optional}{if the selection in \code{selectMethod} does not find
+    a valid method an error is generated, unless \code{optional} is
+    true.  In that case, the value returned is \code{NULL} if no method
+    matches.}
+  \item{mlist, fdef, useInherited, verbose, doCache}{optional arguments
+    to  \code{getMethod} and \code{selectMethod} for internal use.  Avoid
+    these: some will work as expected and others will not, and none of
+    them is required for normal use of the functions.}
 }
 \details{
-  The \code{signature} argument specifies classes, corresponding to formal arguments of the generic function; to be precise, to the \code{signature} slot of the generic function object.
-    The argument may be a vector of strings identifying
-    classes, and may be named or not.  Names, if supplied, match the
-    names of those formal arguments included in the signature of the
-    generic.  That signature is normally all the arguments except
-    \dots.  However, generic functions can be specified with only a
-    subset of the arguments permitted, or with the signature taking
-    the arguments in a different order.
-
-    It's a good idea to name the arguments in the signature to avoid
-    confusion, if you're dealing with a generic that does something
-    special with its signature.   In any case, the elements of the
-    signature are matched to the formal signature by the same rules
-    used in matching arguments in function calls (see
-    \code{\link{match.call}}).
-
-    The strings in the signature may be class names, \code{"missing"}
-    or \code{"ANY"}.  See \link{Methods} for the meaning of these in
-    method selection.  Arguments not supplied in the signature
-    implicitly correspond to class \code{"ANY"}; in particular, giving
-    an empty signature means to look for the default method.
+  The \code{signature} argument specifies classes, corresponding to
+  formal arguments of the generic function; to be precise, to the
+  \code{signature} slot of the generic function object.  The argument
+  may be a vector of strings identifying classes, and may be named or
+  not.  Names, if supplied, match the names of those formal arguments
+  included in the signature of the generic.  That signature is normally
+  all the arguments except \dots.  However, generic functions can be
+  specified with only a subset of the arguments permitted, or with the
+  signature taking the arguments in a different order.
+
+  It's a good idea to name the arguments in the signature to avoid
+  confusion, if you're dealing with a generic that does something
+  special with its signature.  In any case, the elements of the
+  signature are matched to the formal signature by the same rules used
+  in matching arguments in function calls (see
+  \code{\link{match.call}}).
+
+  The strings in the signature may be class names, \code{"missing"} or
+  \code{"ANY"}.  See \link{Methods} for the meaning of these in method
+  selection.  Arguments not supplied in the signature implicitly
+  correspond to class \code{"ANY"}; in particular, giving an empty
+  signature means to look for the default method.
 
   A call to \code{getMethod} returns the method for a particular
   function and signature.  As with other \code{get} functions,
Binary files R-release-orig/src/library/parallel/inst/doc/parallel.pdf and R-patched/src/library/parallel/inst/doc/parallel.pdf differ
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/parallel/tests/multicore2.Rout.save R-patched/src/library/parallel/tests/multicore2.Rout.save
--- R-release-orig/src/library/parallel/tests/multicore2.Rout.save	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/src/library/parallel/tests/multicore2.Rout.save	2016-07-21 17:45:11.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -45,4 +45,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.146   0.026   0.262 
+  0.113   0.013   0.212 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/parallel/tests/snow2.Rout.save R-patched/src/library/parallel/tests/snow2.Rout.save
--- R-release-orig/src/library/parallel/tests/snow2.Rout.save	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/src/library/parallel/tests/snow2.Rout.save	2016-07-21 17:45:11.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -50,4 +50,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.151   0.024   0.571 
+  0.151   0.022   0.533 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/stats/tests/glm.Rout.save R-patched/src/library/stats/tests/glm.Rout.save
--- R-release-orig/src/library/stats/tests/glm.Rout.save	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/src/library/stats/tests/glm.Rout.save	2016-07-21 17:45:11.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -29,4 +29,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.112   0.020   0.123 
+  0.143   0.019   0.149 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/stats/tests/ks-test.Rout.save R-patched/src/library/stats/tests/ks-test.Rout.save
--- R-release-orig/src/library/stats/tests/ks-test.Rout.save	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/src/library/stats/tests/ks-test.Rout.save	2016-07-21 17:45:12.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -114,4 +114,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.144   0.017   0.148 
+  0.133   0.026   0.148 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/stats/tests/nls.Rout.save R-patched/src/library/stats/tests/nls.Rout.save
--- R-release-orig/src/library/stats/tests/nls.Rout.save	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/src/library/stats/tests/nls.Rout.save	2016-07-21 17:45:12.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -703,4 +703,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  1.110   0.022   1.120 
+  1.217   0.026   1.230 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/stats/tests/simulate.Rout.save R-patched/src/library/stats/tests/simulate.Rout.save
--- R-release-orig/src/library/stats/tests/simulate.Rout.save	2016-06-15 00:15:06.000000000 +0200
+++ R-patched/src/library/stats/tests/simulate.Rout.save	2016-07-21 17:45:12.000000000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -408,4 +408,4 @@
 > 
 > proc.time()
    user  system elapsed 
-  0.202   0.021   0.213 
+  0.204   0.022   0.216 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/tools/R/QC.R R-patched/src/library/tools/R/QC.R
--- R-release-orig/src/library/tools/R/QC.R	2016-04-19 00:15:06.000000000 +0200
+++ R-patched/src/library/tools/R/QC.R	2016-06-23 17:45:11.000000000 +0200
@@ -3747,7 +3747,8 @@
     mfile <- paths[1L]
     make <- Sys.getenv("MAKE")
     if(make == "") make <- "make"
-    command <- sprintf("%s -f %s -f %s -f %s",
+    ## needs a target to avoid targets in src/Makevars
+    command <- sprintf("%s -f %s -f %s -f %s makevars_test",
                        make,
                        shQuote(file.path(R.home("share"), "make",
                                          "check_vars_ini.mk")),
@@ -3792,6 +3793,7 @@
                         "W[^l].*", # -Wl, might just be portable
                         "ansi", "pedantic", "traditional",
                         "f.*", "m.*", "std.*",
+                        "isystem", # gcc and clones
                         "x",
                         "q"),
                       collapse = "|"))
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/tools/R/utils.R R-patched/src/library/tools/R/utils.R
--- R-release-orig/src/library/tools/R/utils.R	2016-04-25 00:15:04.000000000 +0200
+++ R-patched/src/library/tools/R/utils.R	2016-07-21 17:45:12.000000000 +0200
@@ -2071,7 +2071,7 @@
                        tolower(substring(x, 3L)))
             else paste0(toupper(x1), tolower(substring(x, 2L)))
         }
-        xx <- .Call(splitString, x, ' -/"()\n')
+        xx <- .Call(C_splitString, x, ' -/"()\n')
         ## for 'alone' we could insist on that exact capitalization
         alone <- xx %in% c(alone, either)
         alone <- alone | grepl("^'.*'$", xx)
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/tools/src/gramLatex.c R-patched/src/library/tools/src/gramLatex.c
--- R-release-orig/src/library/tools/src/gramLatex.c	2016-06-21 00:19:26.000000000 +0200
+++ R-patched/src/library/tools/src/gramLatex.c	2016-07-20 17:45:09.000000000 +0200
@@ -2709,7 +2709,7 @@
 
 /* "do_parseLatex" 
 
- .External2(CC_parseLatex, file, srcfile, verbose, basename, warningCalls)
+ .External2(C_parseLatex, file, srcfile, verbose, basename, warningCalls)
  If there is text then that is read and the other arguments are ignored.
 */
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/tools/src/gramLatex.y R-patched/src/library/tools/src/gramLatex.y
--- R-release-orig/src/library/tools/src/gramLatex.y	2016-03-17 00:03:30.000000000 +0100
+++ R-patched/src/library/tools/src/gramLatex.y	2016-07-20 17:45:09.000000000 +0200
@@ -900,7 +900,7 @@
 
 /* "do_parseLatex" 
 
- .External2(CC_parseLatex, file, srcfile, verbose, basename, warningCalls)
+ .External2(C_parseLatex, file, srcfile, verbose, basename, warningCalls)
  If there is text then that is read and the other arguments are ignored.
 */
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/tools/src/init.c R-patched/src/library/tools/src/init.c
--- R-release-orig/src/library/tools/src/init.c	2016-03-17 00:03:30.000000000 +0100
+++ R-patched/src/library/tools/src/init.c	2016-07-21 17:45:12.000000000 +0200
@@ -51,6 +51,7 @@
     CALLDEF(stopHTTPD, 0),
     CALLDEF(C_deparseRd, 2),
     CALLDEF(splitString, 2),
+    {"C_splitString", (DL_FUNC) &splitString, 2},
 
     {NULL, NULL, 0}
 };
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/utils/NAMESPACE R-patched/src/library/utils/NAMESPACE
--- R-release-orig/src/library/utils/NAMESPACE	2016-03-25 00:15:12.000000000 +0100
+++ R-patched/src/library/utils/NAMESPACE	2016-06-29 02:41:11.615844000 +0200
@@ -20,7 +20,7 @@
        flush.console, formatOL, formatUL, getAnywhere, getCRANmirrors,
        getFromNamespace, getParseData, getParseText, getS3method,
        getSrcDirectory, getSrcFilename, getSrcLocation, getSrcref, # getRcode,
-       glob2rx, globalVariables, head, head.matrix, help,
+       glob2rx, globalVariables, hasName, head, head.matrix, help,
        help.request, help.search, help.start, history,
        install.packages, installed.packages, is.relistable, isS3method,
        limitedLabels, loadhistory, localeToCharset, ls.str, lsf.str,
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/utils/R/hasName.R R-patched/src/library/utils/R/hasName.R
--- R-release-orig/src/library/utils/R/hasName.R	1970-01-01 01:00:00.000000000 +0100
+++ R-patched/src/library/utils/R/hasName.R	2016-07-15 02:41:26.000000000 +0200
@@ -0,0 +1,21 @@
+#  File src/library/utils/R/hasName.R
+#  Part of the R package, https://www.R-project.org
+#
+#  Copyright (C) 2016 The R Core Team
+#
+#  This program is free software; you can redistribute it and/or modify
+#  it under the terms of the GNU General Public License as published by
+#  the Free Software Foundation; either version 2 of the License, or
+#  (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+#  GNU General Public License for more details.
+#
+#  A copy of the GNU General Public License is available at
+#  https://www.R-project.org/Licenses/
+
+# "hasName(x, name)" is defined to be the same as "name %in% names(x)"
+
+hasName <- function(x, name) match(name, names(x), nomatch = 0L) > 0L
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/utils/R/sessionInfo.R R-patched/src/library/utils/R/sessionInfo.R
--- R-release-orig/src/library/utils/R/sessionInfo.R	2016-06-15 00:15:05.000000000 +0200
+++ R-patched/src/library/utils/R/sessionInfo.R	2016-06-21 17:45:11.000000000 +0200
@@ -56,7 +56,8 @@
                        ver <- sub(".*<string>", "", ver)
                        ver <- sub("</string>$", "", ver)
                        ver1 <- strsplit(ver, ".", fixed = TRUE)[[1L]][2L]
-                       sprintf("OS X %s (%s)", ver,
+                       sprintf("%s %s (%s)",
+                               ifelse(as.numeric(ver1) < 12, "OS X", "macOS"),
                                switch(ver1,
                                       "4" = "Tiger",
                                       "5" = "Leopard",
@@ -67,7 +68,7 @@
                                       "10" = "Yosemite",
                                       "11" = "El Capitan",
                                       "12" = "Sierra",
-                                     "unknown"))
+                                      "unknown"), ver)
                    },
                    "SunOS" = {
                        ver <- system('uname -r', intern = TRUE)
Binary files R-release-orig/src/library/utils/inst/doc/Sweave.pdf and R-patched/src/library/utils/inst/doc/Sweave.pdf differ
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/library/utils/man/hasName.Rd R-patched/src/library/utils/man/hasName.Rd
--- R-release-orig/src/library/utils/man/hasName.Rd	1970-01-01 01:00:00.000000000 +0100
+++ R-patched/src/library/utils/man/hasName.Rd	2016-07-15 02:41:27.000000000 +0200
@@ -0,0 +1,42 @@
+% File src/library/utils/man/hasName.Rd
+% Part of the R package, https://www.R-project.org
+% Copyright 2016 R Core Team
+% Distributed under GPL 2 or later
+
+\name{hasName}
+\alias{hasName}
+\title{Check for name}
+\description{
+  \code{hasName} is a convenient way to test for one or more names
+in an R object.
+}
+\usage{
+hasName(x, name)
+}
+\arguments{
+  \item{x}{Any object.}
+  \item{name}{One or more character values to look for.}
+}
+\value{
+  A logical vector of the same length as \code{name} containing
+  \code{TRUE} if the corresponding entry is in \code{names(x)}.
+}
+\details{
+  \code{hasName(x, name)} is defined to be equivalent to 
+  \code{name \%in\% names(x)}, though it will evaluate slightly more
+  quickly.  It is intended to replace the common idiom
+  \code{!is.null(x$name)}.  The latter can be unreliable due to partial
+  name matching; see the example below.
+}
+\seealso{
+  \code{\link{\%in\%}}, \code{\link{exists}}
+}
+\examples{
+x <- list(abc = 1, def = 2)
+!is.null(x$abc) # correct
+!is.null(x$a)   # this is the wrong test!
+hasName(x, "abc")
+hasName(x, "a")
+}
+\keyword{manip}
+\keyword{logic}
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/main/context.c R-patched/src/main/context.c
--- R-release-orig/src/main/context.c	2016-03-17 00:04:17.000000000 +0100
+++ R-patched/src/main/context.c	2016-07-17 17:45:09.042876000 +0200
@@ -351,6 +351,8 @@
     if (n == 0)
 	return(R_GlobalEnv);
 
+    if (n == NA_INTEGER) error(_("NA argument is invalid"));
+
     if (n > 0)
 	n = framedepth(cptr) - n;
     else
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/main/datetime.c R-patched/src/main/datetime.c
--- R-release-orig/src/main/datetime.c	2016-03-17 00:04:18.000000000 +0100
+++ R-patched/src/main/datetime.c	2016-06-28 17:45:09.549451000 +0200
@@ -759,9 +759,8 @@
     setAttrib(ans, install("tzone"), tzone);
     SEXP nm = getAttrib(x, R_NamesSymbol);
     if(nm != R_NilValue) setAttrib(VECTOR_ELT(ans, 5), R_NamesSymbol, nm);
-    UNPROTECT(6);
-
     if(settz) reset_tz(oldtz);
+    UNPROTECT(6);
     return ans;
 }
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/main/eval.c R-patched/src/main/eval.c
--- R-release-orig/src/main/eval.c	2016-05-21 00:15:07.000000000 +0200
+++ R-patched/src/main/eval.c	2016-07-17 17:45:09.079176000 +0200
@@ -1122,7 +1122,8 @@
 	from the function body.  */
 
     if ((SETJMP(cntxt.cjmpbuf))) {
-	if (R_ReturnedValue == R_RestartToken) {
+	if (! cntxt.jumptarget && /* ignores intermediate jumps for on.exits */
+	    R_ReturnedValue == R_RestartToken) {
 	    cntxt.callflag = CTXT_RETURN;  /* turn restart off */
 	    R_ReturnedValue = R_NilValue;  /* remove restart token */
 	    PROTECT(tmp = eval(body, newrho));
@@ -3919,7 +3920,8 @@
 	if (typex == REALSXP && typey == REALSXP) {			\
 	    double rn1 = vx.dval;					\
 	    double rn2 = vy.dval;					\
-	    if (INT_MIN <= rn1 && INT_MAX >= rn1 &&			\
+	    if (R_FINITE(rn1) && R_FINITE(rn2) &&			\
+		INT_MIN <= rn1 && INT_MAX >= rn1 &&			\
 		INT_MIN <= rn2 && INT_MAX >- rn2 &&			\
 		rn1 == (int) rn1 && rn2 == (int) rn2) {			\
 		SKIP_OP(); /* skip 'call' index */			\
@@ -5158,6 +5160,14 @@
     }
 }
 
+static SEXP markSpecialArgs(SEXP args)
+{
+    SEXP arg;
+    for(arg = args; arg != R_NilValue; arg = CDR(arg))
+	MARK_NOT_MUTABLE(CAR(arg));
+    return args;
+}
+
 static SEXP bcEval(SEXP body, SEXP rho, Rboolean useCache)
 {
   SEXP value = R_NilValue, constants;
@@ -5613,7 +5623,7 @@
 	case SPECIALSXP:
 	  flag = PRIMPRINT(fun);
 	  R_Visible = flag != 1;
-	  value = PRIMFUN(fun) (call, fun, CDR(call), rho);
+	  value = PRIMFUN(fun) (call, fun, markSpecialArgs(CDR(call)), rho);
 	  if (flag < 2) R_Visible = flag != 1;
 	  break;
 	case CLOSXP:
@@ -5665,7 +5675,7 @@
 	}
 	flag = PRIMPRINT(fun);
 	R_Visible = flag != 1;
-	value = PRIMFUN(fun) (call, fun, CDR(call), rho);
+	value = PRIMFUN(fun) (call, fun, markSpecialArgs(CDR(call)), rho);
 	if (flag < 2) R_Visible = flag != 1;
 	vmaxset(vmax);
 	BCNPUSH(value);
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/main/gevents.c R-patched/src/main/gevents.c
--- R-release-orig/src/main/gevents.c	2016-03-17 00:04:15.000000000 +0100
+++ R-patched/src/main/gevents.c	2016-07-17 17:45:09.104785000 +0200
@@ -90,7 +90,10 @@
 
     checkArity(op, args);
 
-    devnum = INTEGER(CAR(args))[0] - 1;
+    devnum = INTEGER(CAR(args))[0];
+    if(devnum == NA_INTEGER)
+	error(_("invalid graphical device number"));
+    devnum--;
     if(devnum < 1 || devnum > R_MaxDevices)
 	error(_("invalid graphical device number"));
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/main/platform.c R-patched/src/main/platform.c
--- R-release-orig/src/main/platform.c	2016-03-17 00:04:17.000000000 +0100
+++ R-patched/src/main/platform.c	2016-06-21 17:45:13.389789000 +0200
@@ -1,7 +1,7 @@
 /*
  *  R : A Computer Language for Statistical Data Analysis
  *  Copyright (C) 1995, 1996  Robert Gentleman and Ross Ihaka
- *  Copyright (C) 1998--2015 The R Core Team
+ *  Copyright (C) 1998--2016 The R Core Team
  *
  *  This program is free software; you can redistribute it and/or modify
  *  it under the terms of the GNU General Public License as published by
@@ -2919,12 +2919,20 @@
 # include <gnu/libc-version.h>
 #endif
 
+#ifdef HAVE_LIBREADLINE
+# ifdef HAVE_READLINE_READLINE_H
+#  include <readline/readline.h>
+# else
+extern const char *rl_library_version;
+# endif
+#endif
+
 SEXP attribute_hidden
 do_eSoftVersion(SEXP call, SEXP op, SEXP args, SEXP rho)
 {
     checkArity(op, args);
-    SEXP ans = PROTECT(allocVector(STRSXP, 7));
-    SEXP nms = PROTECT(allocVector(STRSXP, 7));
+    SEXP ans = PROTECT(allocVector(STRSXP, 8));
+    SEXP nms = PROTECT(allocVector(STRSXP, 8));
     setAttrib(ans, R_NamesSymbol, nms);
     unsigned int i = 0;
     char p[256];
@@ -2967,6 +2975,12 @@
 #endif
     SET_STRING_ELT(ans, i, mkChar(p));
     SET_STRING_ELT(nms, i++, mkChar("iconv"));
+#ifdef HAVE_LIBREADLINE
+    SET_STRING_ELT(ans, i, mkChar(rl_library_version));
+#else
+    SET_STRING_ELT(ans, i, mkChar(""));
+#endif
+    SET_STRING_ELT(nms, i++, mkChar("readline"));
     UNPROTECT(2);
     return ans;
 }
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/src/unix/sys-std.c R-patched/src/unix/sys-std.c
--- R-release-orig/src/unix/sys-std.c	2016-06-10 00:15:06.000000000 +0200
+++ R-patched/src/unix/sys-std.c	2016-06-21 17:45:13.591444000 +0200
@@ -417,7 +417,6 @@
 }
 
 
-
 #ifdef HAVE_LIBREADLINE
 
 # ifdef HAVE_READLINE_READLINE_H
@@ -427,6 +426,15 @@
 #  if !defined (_RL_FUNCTION_TYPEDEF)
 typedef void rl_vcpfunc_t (char *);
 #  endif /* _RL_FUNCTION_TYPEDEF */
+
+# if defined(RL_READLINE_VERSION) && RL_READLINE_VERSION >= 0x0603
+/* readline 6.3's rl_callback_handler_install() no longer installs
+   signal handlers, so as from that version we need an explicit
+   one. (PR#16604)  (This could have been controlled in earlier versions
+   by setting rl_catch_sigwinch.)
+ */
+#  define NEED_INT_HANDLER
+# endif
 # else
 typedef void rl_vcpfunc_t (char *);
 extern void rl_callback_handler_install(const char *, rl_vcpfunc_t *);
@@ -513,6 +521,15 @@
   rl_vcpfunc_t *fun[MAX_READLINE_NESTING];
 } ReadlineStack = {-1, MAX_READLINE_NESTING - 1};
 
+#ifdef NEED_INT_HANDLER
+static volatile Rboolean caught_sigwinch = FALSE;
+
+static RETSIGTYPE
+R_readline_sigwinch_handler(int sig)
+{
+    caught_sigwinch = TRUE;
+}
+#endif
 
 /*
   Registers the specified routine and prompt with readline
@@ -527,6 +544,11 @@
      ReadlineStack.fun[++ReadlineStack.current] = f;
 
    rl_callback_handler_install(prompt, f);
+
+#ifdef NEED_INT_HANDLER
+   signal(SIGWINCH, R_readline_sigwinch_handler);
+#endif
+
    /* flush stdout in case readline wrote the prompt, but didn't flush
       stdout to make it visible. (needed for Apple's readline emulation). */
    fflush(stdout);
@@ -534,7 +556,7 @@
 
 #if defined(RL_READLINE_VERSION) && RL_READLINE_VERSION >= 0x0600
 /*
-  Fix for PR#16603.
+  Fix for PR#16603, for readline >= 6.0.
 
   The readline interface is somewhat messy. readline contains the
   function rl_free_line_state(), which its internal SIGINT handler
@@ -716,9 +738,8 @@
     /* Tell the completer that we want a crack first. */
     rl_attempted_completion_function = R_custom_completion;
 
-    /* Disable sorting of possible completions; only readline >= 6 */
+    /* Disable sorting of possible completions; only readline >= 6.0 */
 #if RL_READLINE_VERSION >= 0x0600
-    /* if (rl_readline_version >= 0x0600) */
     rl_sort_completion_matches = 0;
 #endif
 
@@ -954,6 +975,14 @@
 	    if (Rg_wait_usec > 0 && (wt < 0 || wt > Rg_wait_usec))
 		wt = Rg_wait_usec;
 	    what = R_checkActivityEx(wt, 0, handleInterrupt);
+#ifdef NEED_INT_HANDLER
+            if (UsingReadline && caught_sigwinch) {
+		caught_sigwinch = FALSE;
+		// iintroduced in readline 4.0: only used for >= 6.3
+		rl_resize_terminal();
+            }
+#endif
+
 	    /* This is slightly clumsy. We have advertised the
 	     * convention that R_wait_usec == 0 means "wait forever",
 	     * but we also need to enable R_checkActivity to return
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/datasets-Ex.Rout.save R-patched/tests/Examples/datasets-Ex.Rout.save
--- R-release-orig/tests/Examples/datasets-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/datasets-Ex.Rout.save	2016-07-21 17:45:14.083959000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -3458,7 +3458,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  1.27 0.028 1.299 0 0 
+Time elapsed:  1.273 0.025 1.298 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/grDevices-Ex.Rout.save R-patched/tests/Examples/grDevices-Ex.Rout.save
--- R-release-orig/tests/Examples/grDevices-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/grDevices-Ex.Rout.save	2016-07-21 17:45:14.126370000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -3766,7 +3766,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  7.675 0.138 7.813 0 0 
+Time elapsed:  7.906 0.141 8.047 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/graphics-Ex.Rout.save R-patched/tests/Examples/graphics-Ex.Rout.save
--- R-release-orig/tests/Examples/graphics-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/graphics-Ex.Rout.save	2016-07-21 17:45:14.185799000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -2066,7 +2066,7 @@
 > points.default # to see how it calls "plot.xy(xy.coords(x, y), ...)"
 function (x, y = NULL, type = "p", ...) 
 plot.xy(xy.coords(x, y), type = type, ...)
-<bytecode: 0x29093c8>
+<bytecode: 0x1e47ad8>
 <environment: namespace:graphics>
 > 
 > 
@@ -2936,7 +2936,7 @@
     u <- par("usr")
     xy * c(u[2L] - u[1L], u[4L] - u[3L])/par("pin")
 }
-<bytecode: 0x3ee75b0>
+<bytecode: 0x2093d88>
 <environment: namespace:graphics>
 > 
 > ## plot labels offset 0.12 inches to the right
@@ -3016,7 +3016,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  1.809 0.027 1.836 0 0 
+Time elapsed:  1.817 0.027 1.845 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/grid-Ex.Rout.save R-patched/tests/Examples/grid-Ex.Rout.save
--- R-release-orig/tests/Examples/grid-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/grid-Ex.Rout.save	2016-07-21 17:45:14.224206000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -2069,7 +2069,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  1.098 0.007 1.105 0 0 
+Time elapsed:  1.087 0.012 1.099 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/splines-Ex.Rout.save R-patched/tests/Examples/splines-Ex.Rout.save
--- R-release-orig/tests/Examples/splines-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/splines-Ex.Rout.save	2016-07-21 17:45:14.263476000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -582,7 +582,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  0.153 0.005 0.159 0 0 
+Time elapsed:  0.104 0.002 0.106 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/stats-Ex.Rout.save R-patched/tests/Examples/stats-Ex.Rout.save
--- R-release-orig/tests/Examples/stats-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/stats-Ex.Rout.save	2016-07-21 17:45:14.323277321 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -3792,7 +3792,7 @@
 > 
 > asOneSidedFormula("age")
 ~age
-<environment: 0x34aa138>
+<environment: 0x2b084e8>
 > asOneSidedFormula(~ age)
 ~age
 > 
@@ -4755,7 +4755,7 @@
     r[cbind(1L:p, 1L:p)] <- 1
     r
 }
-<bytecode: 0x370b258>
+<bytecode: 0x28a5150>
 <environment: namespace:stats>
 > stopifnot(all.equal(Cl, cov2cor(cov(longley))),
 +           all.equal(cor(longley, method = "kendall"),
@@ -7285,7 +7285,7 @@
 > environment(as.formula("y ~ x"))
 <environment: R_GlobalEnv>
 > environment(as.formula("y ~ x", env = new.env()))
-<environment: 0x34f71e0>
+<environment: 0x2678478>
 > 
 > 
 > ## Create a formula for a model with a large number of variables:
@@ -11766,14 +11766,14 @@
 $linkfun
 function (mu) 
 mu^lambda
-<bytecode: 0x6f42160>
-<environment: 0x6f4ab40>
+<bytecode: 0x60d0f70>
+<environment: 0x60d89b0>
 
 $linkinv
 function (eta) 
 pmax(eta^(1/lambda), .Machine$double.eps)
-<bytecode: 0x6f42fb0>
-<environment: 0x6f4ab40>
+<bytecode: 0x60d0e58>
+<environment: 0x60d89b0>
 
 > 
 > 
@@ -15771,8 +15771,8 @@
 > unclass(sfun0)
 function (v) 
 .approxfun(x, y, v, method, yleft, yright, f)
-<bytecode: 0x2d6cec8>
-<environment: 0x7356b68>
+<bytecode: 0x1f0ba48>
+<environment: 0x64b29a0>
 attr(,"call")
 stepfun(1:3, y0, f = 0)
 > ls(envir = environment(sfun0))
@@ -18049,7 +18049,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  4.26 0.087 4.348 0 0 
+Time elapsed:  4.326 0.102 4.428 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/stats4-Ex.Rout.save R-patched/tests/Examples/stats4-Ex.Rout.save
--- R-release-orig/tests/Examples/stats4-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/stats4-Ex.Rout.save	2016-07-21 17:45:14.373851000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -184,7 +184,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  0.289 0.003 0.292 0 0 
+Time elapsed:  0.291 0.001 0.292 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/Examples/tools-Ex.Rout.save R-patched/tests/Examples/tools-Ex.Rout.save
--- R-release-orig/tests/Examples/tools-Ex.Rout.save	2016-06-15 00:15:07.000000000 +0200
+++ R-patched/tests/Examples/tools-Ex.Rout.save	2016-07-21 17:45:14.409688000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -188,7 +188,7 @@
 > bibstyle("unsorted", sortKeys = function(refs) seq_along(refs),
 +     fmtPrefix = function(paper) paste0("[", paper$.index, "]"),
 +        .init = TRUE)
-<environment: 0x3582e28>
+<environment: 0x356cd68>
 > print(refs, .bibstyle = "unsorted")
 [1] R Core Team (2013). _R: A Language and Environment for Statistical
 Computing_. R Foundation for Statistical Computing, Vienna, Austria.
@@ -1015,7 +1015,7 @@
 > ###
 > options(digits = 7L)
 > base::cat("Time elapsed: ", proc.time() - base::get("ptime", pos = 'CheckExEnv'),"\n")
-Time elapsed:  0.191 0.004 0.197 0 0 
+Time elapsed:  0.184 0.007 0.191 0 0 
 > grDevices::dev.off()
 null device 
           1 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/R-intro.Rout.save R-patched/tests/R-intro.Rout.save
--- R-release-orig/tests/R-intro.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/R-intro.Rout.save	2016-07-21 17:45:19.086865000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/any-all.Rout.save R-patched/tests/any-all.Rout.save
--- R-release-orig/tests/any-all.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/any-all.Rout.save	2016-07-21 17:45:19.129762000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/arith-true.Rout.save R-patched/tests/arith-true.Rout.save
--- R-release-orig/tests/arith-true.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/arith-true.Rout.save	2016-07-21 17:45:19.176004000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -323,5 +323,5 @@
 > 
 > ## Last Line:
 > cat('Time elapsed: ', proc.time() - .proctime00,'\n')
-Time elapsed:  0.138 0.002 0.139 0 0 
+Time elapsed:  0.132 0 0.132 0 0 
 > 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/arith.Rout.save R-patched/tests/arith.Rout.save
--- R-release-orig/tests/arith.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/arith.Rout.save	2016-07-21 17:45:19.202067000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/complex.Rout.save R-patched/tests/complex.Rout.save
--- R-release-orig/tests/complex.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/complex.Rout.save	2016-07-21 17:45:19.233701000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/d-p-q-r-tests.Rout.save R-patched/tests/d-p-q-r-tests.Rout.save
--- R-release-orig/tests/d-p-q-r-tests.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/d-p-q-r-tests.Rout.save	2016-07-21 17:45:19.273002000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -1388,5 +1388,5 @@
 > 
 > 
 > cat("Time elapsed: ", proc.time() - .ptime,"\n")
-Time elapsed:  0.954 0.014 0.967 0 0 
+Time elapsed:  0.901 0.006 0.906 0 0 
 > 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/datasets.Rout.save R-patched/tests/datasets.Rout.save
--- R-release-orig/tests/datasets.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/datasets.Rout.save	2016-07-21 17:45:19.299116000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/datetime.Rout.save R-patched/tests/datetime.Rout.save
--- R-release-orig/tests/datetime.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/datetime.Rout.save	2016-07-21 17:45:19.324154000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/datetime2.Rout.save R-patched/tests/datetime2.Rout.save
--- R-release-orig/tests/datetime2.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/datetime2.Rout.save	2016-07-21 17:45:19.360411000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/demos.Rout.save R-patched/tests/demos.Rout.save
--- R-release-orig/tests/demos.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/demos.Rout.save	2016-07-21 17:45:19.419453000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -1510,5 +1510,5 @@
 > par(op)
 > 
 > cat("Time elapsed: ", proc.time() - .ptime, "\n")
-Time elapsed:  0.977 0.026 1.004 0 0 
+Time elapsed:  1.044 0.017 1.061 0 0 
 > 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/eval-etc.Rout.save R-patched/tests/eval-etc.Rout.save
--- R-release-orig/tests/eval-etc.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/eval-etc.Rout.save	2016-07-21 17:45:19.465466000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/iec60559.Rout.save R-patched/tests/iec60559.Rout.save
--- R-release-orig/tests/iec60559.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/iec60559.Rout.save	2016-07-21 17:45:19.526075000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/internet.Rout.save R-patched/tests/internet.Rout.save
--- R-release-orig/tests/internet.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/internet.Rout.save	2016-07-21 17:45:19.576965000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -23,7 +23,7 @@
 > 
 > # test do_download.
 > nrow(available.packages(contrib.url("http://cran.r-project.org")))
-[1] 8572
+[1] 8790
 > 
 > # test url connections on http
 > zz <- url("http://cran.r-project.org/")
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/isas-tests.Rout.save R-patched/tests/isas-tests.Rout.save
--- R-release-orig/tests/isas-tests.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/isas-tests.Rout.save	2016-07-21 17:45:19.619363000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -2568,6 +2568,6 @@
 > f <- try(as.ts( x ), silent = TRUE)
 >   if(!inherits(f, 'try-error')) report(identical(f, as.ts( f )))
 > cat('Time elapsed: ', proc.time() - .proctime00,'\n')
-Time elapsed:  0.129 0.009 0.136 0 0 
+Time elapsed:  0.131 0.01 0.142 0 0 
 > 
 > 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/lapack.Rout.save R-patched/tests/lapack.Rout.save
--- R-release-orig/tests/lapack.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/lapack.Rout.save	2016-07-21 17:45:19.665962000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/lm-tests.Rout.save R-patched/tests/lm-tests.Rout.save
--- R-release-orig/tests/lm-tests.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/lm-tests.Rout.save	2016-07-21 17:45:19.719331000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/method-dispatch.Rout.save R-patched/tests/method-dispatch.Rout.save
--- R-release-orig/tests/method-dispatch.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/method-dispatch.Rout.save	2016-07-21 17:45:19.752672000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/ok-errors.Rout.save R-patched/tests/ok-errors.Rout.save
--- R-release-orig/tests/ok-errors.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/ok-errors.Rout.save	2016-07-21 17:45:19.796050000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/p-r-random-tests.Rout.save R-patched/tests/p-r-random-tests.Rout.save
--- R-release-orig/tests/p-r-random-tests.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/p-r-random-tests.Rout.save	2016-07-21 17:45:19.850561000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -278,6 +278,6 @@
 > 
 > 
 > cat('Time elapsed: ', proc.time() - .proctime00,'\n')
-Time elapsed:  1.202 0.003 1.205 0 0 
+Time elapsed:  1.135 0.002 1.136 0 0 
 > 
 > 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/print-tests.Rout.save R-patched/tests/print-tests.Rout.save
--- R-release-orig/tests/print-tests.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/print-tests.Rout.save	2016-07-21 17:45:19.898533000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-IO.Rout.save R-patched/tests/reg-IO.Rout.save
--- R-release-orig/tests/reg-IO.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/reg-IO.Rout.save	2016-07-21 17:45:19.929315000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-IO2.Rout.save R-patched/tests/reg-IO2.Rout.save
--- R-release-orig/tests/reg-IO2.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/reg-IO2.Rout.save	2016-07-21 17:45:20.010045000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-S4.Rout.save R-patched/tests/reg-S4.Rout.save
--- R-release-orig/tests/reg-S4.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/reg-S4.Rout.save	2016-07-21 17:45:20.094644000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-examples3.Rout.save R-patched/tests/reg-examples3.Rout.save
--- R-release-orig/tests/reg-examples3.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/reg-examples3.Rout.save	2016-07-21 17:45:20.229513000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-packages.R R-patched/tests/reg-packages.R
--- R-release-orig/tests/reg-packages.R	2016-03-17 00:04:34.000000000 +0100
+++ R-patched/tests/reg-packages.R	2016-07-13 12:28:49.038920000 +0200
@@ -141,12 +141,19 @@
     ## Both exNSS4 and Matrix define "atomicVector" *the same*,
     ## but  'exNSS4'  has it extended - and hence *both* are registered in cache -> "conflicts"
     requireNamespace("exNSS4", lib= "myLib")
+    ## Found in cache, since there is only one definition.
+    ## Might confuse users.
+    stopifnot(isVirtualClass(getClass("atomicVector")))
     requireNamespace("Matrix", lib= .Library)
-    tools::assertCondition( ## condition, because this *still* uses message():
+    ## Throws an error, because there is ambiguity in the cache,
+    ## and the dynamic search will not find anything, since the packages
+    ## are not attached.
+    tools::assertCondition(
         acl <- getClass("atomicVector")
-    ) ## gave an Error: “atomicVector” is not a defined class
-    ##   ... because it was found non-uniquely
-    stopifnot(is(acl, "classRepresentation"), isVirtualClass(acl))
+        )
+    ## Once Matrix is attached, we find a unique definition.
+    library(Matrix)
+    stopifnot(isVirtualClass(getClass("atomicVector")))
 }
 
 ## clean up
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-plot.Rout.save R-patched/tests/reg-plot.Rout.save
--- R-release-orig/tests/reg-plot.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/reg-plot.Rout.save	2016-07-21 17:45:20.266985000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-tests-2.Rout.save R-patched/tests/reg-tests-2.Rout.save
--- R-release-orig/tests/reg-tests-2.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/reg-tests-2.Rout.save	2016-07-21 17:45:20.315428000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
@@ -6405,15 +6405,15 @@
 > print.function
 function (x, ...)  
  - attr(*, "srcref")=Class 'srcref'  atomic [1:8] 1 19 1 63 19 63 1 1
-  .. ..- attr(*, "srcfile")=Classes 'srcfilecopy', 'srcfile' <environment: 0x4133960> 
+  .. ..- attr(*, "srcfile")=Classes 'srcfilecopy', 'srcfile' <environment: 0x3a990a0> 
 > f
 function ()  
  - attr(*, "srcref")=Class 'srcref'  atomic [1:8] 1 17 1 28 17 28 1 1
-  .. ..- attr(*, "srcfile")=Classes 'srcfilecopy', 'srcfile' <environment: 0x4125bf8> 
+  .. ..- attr(*, "srcfile")=Classes 'srcfilecopy', 'srcfile' <environment: 0x3a8f108> 
  - attr(*, "note")= chr "just a note"
  - attr(*, "yada")=function ()  
   ..- attr(*, "srcref")=Class 'srcref'  atomic [1:8] 2 24 2 48 24 48 2 2
-  .. .. ..- attr(*, "srcfile")=Classes 'srcfilecopy', 'srcfile' <environment: 0x4125bf8> 
+  .. .. ..- attr(*, "srcfile")=Classes 'srcfilecopy', 'srcfile' <environment: 0x3a8f108> 
 > rm(print.function)
 > ## auto-printing and printing differed up to R 2.9.x
 > 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/reg-tests-3.Rout.save R-patched/tests/reg-tests-3.Rout.save
--- R-release-orig/tests/reg-tests-3.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/reg-tests-3.Rout.save	2016-07-21 17:45:20.350647000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tests/simple-true.Rout.save R-patched/tests/simple-true.Rout.save
--- R-release-orig/tests/simple-true.Rout.save	2016-06-15 00:15:10.000000000 +0200
+++ R-patched/tests/simple-true.Rout.save	2016-07-21 17:45:20.383286000 +0200
@@ -1,5 +1,5 @@
 
-R version 3.3.1 RC (2016-06-14 r70774) -- "Bug in Your Hair"
+R version 3.3.1 Patched (2016-07-21 r70950) -- "Bug in Your Hair"
 Copyright (C) 2016 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tools/config.guess R-patched/tools/config.guess
--- R-release-orig/tools/config.guess	2016-01-08 00:15:04.000000000 +0100
+++ R-patched/tools/config.guess	2016-06-21 17:45:14.057743000 +0200
@@ -2,7 +2,7 @@
 # Attempt to guess a canonical system name.
 #   Copyright 1992-2016 Free Software Foundation, Inc.
 
-timestamp='2016-01-01'
+timestamp='2016-06-15'
 
 # This file is free software; you can redistribute it and/or modify it
 # under the terms of the GNU General Public License as published by
@@ -186,9 +186,12 @@
 	    *) machine=${UNAME_MACHINE_ARCH}-unknown ;;
 	esac
 	# The Operating System including object format, if it has switched
-	# to ELF recently, or will in the future.
+	# to ELF recently (or will in the future) and ABI.
 	case "${UNAME_MACHINE_ARCH}" in
-	    arm*|earm*|i386|m68k|ns32k|sh3*|sparc|vax)
+	    earm*)
+		os=netbsdelf
+		;;
+	    arm*|i386|m68k|ns32k|sh3*|sparc|vax)
 		eval $set_cc_for_build
 		if echo __ELF__ | $CC_FOR_BUILD -E - 2>/dev/null \
 			| grep -q __ELF__
@@ -237,6 +240,10 @@
 	UNAME_MACHINE_ARCH=`arch | sed 's/OpenBSD.//'`
 	echo ${UNAME_MACHINE_ARCH}-unknown-openbsd${UNAME_RELEASE}
 	exit ;;
+    *:LibertyBSD:*:*)
+	UNAME_MACHINE_ARCH=`arch | sed 's/^.*BSD\.//'`
+	echo ${UNAME_MACHINE_ARCH}-unknown-libertybsd${UNAME_RELEASE}
+	exit ;;
     *:ekkoBSD:*:*)
 	echo ${UNAME_MACHINE}-unknown-ekkobsd${UNAME_RELEASE}
 	exit ;;
@@ -268,42 +275,42 @@
 	ALPHA_CPU_TYPE=`/usr/sbin/psrinfo -v | sed -n -e 's/^  The alpha $.*$ processor.*$/\1/p' | head -n 1`
 	case "$ALPHA_CPU_TYPE" in
 	    "EV4 (21064)")
-		UNAME_MACHINE="alpha" ;;
+		UNAME_MACHINE=alpha ;;
 	    "EV4.5 (21064)")
-		UNAME_MACHINE="alpha" ;;
+		UNAME_MACHINE=alpha ;;
 	    "LCA4 (21066/21068)")
-		UNAME_MACHINE="alpha" ;;
+		UNAME_MACHINE=alpha ;;
 	    "EV5 (21164)")
-		UNAME_MACHINE="alphaev5" ;;
+		UNAME_MACHINE=alphaev5 ;;
 	    "EV5.6 (21164A)")
-		UNAME_MACHINE="alphaev56" ;;
+		UNAME_MACHINE=alphaev56 ;;
 	    "EV5.6 (21164PC)")
-		UNAME_MACHINE="alphapca56" ;;
+		UNAME_MACHINE=alphapca56 ;;
 	    "EV5.7 (21164PC)")
-		UNAME_MACHINE="alphapca57" ;;
+		UNAME_MACHINE=alphapca57 ;;
 	    "EV6 (21264)")
-		UNAME_MACHINE="alphaev6" ;;
+		UNAME_MACHINE=alphaev6 ;;
 	    "EV6.7 (21264A)")
-		UNAME_MACHINE="alphaev67" ;;
+		UNAME_MACHINE=alphaev67 ;;
 	    "EV6.8CB (21264C)")
-		UNAME_MACHINE="alphaev68" ;;
+		UNAME_MACHINE=alphaev68 ;;
 	    "EV6.8AL (21264B)")
-		UNAME_MACHINE="alphaev68" ;;
+		UNAME_MACHINE=alphaev68 ;;
 	    "EV6.8CX (21264D)")
-		UNAME_MACHINE="alphaev68" ;;
+		UNAME_MACHINE=alphaev68 ;;
 	    "EV6.9A (21264/EV69A)")
-		UNAME_MACHINE="alphaev69" ;;
+		UNAME_MACHINE=alphaev69 ;;
 	    "EV7 (21364)")
-		UNAME_MACHINE="alphaev7" ;;
+		UNAME_MACHINE=alphaev7 ;;
 	    "EV7.9 (21364A)")
-		UNAME_MACHINE="alphaev79" ;;
+		UNAME_MACHINE=alphaev79 ;;
 	esac
 	# A Pn.n version is a patched version.
 	# A Vn.n version is a released version.
 	# A Tn.n version is a released field test version.
 	# A Xn.n version is an unreleased experimental baselevel.
 	# 1.2 uses "1.2" for uname -r.
-	echo ${UNAME_MACHINE}-dec-osf`echo ${UNAME_RELEASE} | sed -e 's/^[PVTX]//' | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz'`
+	echo ${UNAME_MACHINE}-dec-osf`echo ${UNAME_RELEASE} | sed -e 's/^[PVTX]//' | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz`
 	# Reset EXIT trap before exiting to avoid spurious non-zero exit code.
 	exitcode=$?
 	trap '' 0
@@ -376,16 +383,16 @@
 	exit ;;
     i86pc:SunOS:5.*:* | i86xen:SunOS:5.*:*)
 	eval $set_cc_for_build
-	SUN_ARCH="i386"
+	SUN_ARCH=i386
 	# If there is a compiler, see if it is configured for 64-bit objects.
 	# Note that the Sun cc does not turn __LP64__ into 1 like gcc does.
 	# This test works for both compilers.
-	if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then
+	if [ "$CC_FOR_BUILD" != no_compiler_found ]; then
 	    if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \
-		(CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \
+		(CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \
 		grep IS_64BIT_ARCH >/dev/null
 	    then
-		SUN_ARCH="x86_64"
+		SUN_ARCH=x86_64
 	    fi
 	fi
 	echo ${SUN_ARCH}-pc-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
@@ -410,7 +417,7 @@
 	exit ;;
     sun*:*:4.2BSD:*)
 	UNAME_RELEASE=`(sed 1q /etc/motd | awk '{print substr($5,1,3)}') 2>/dev/null`
-	test "x${UNAME_RELEASE}" = "x" && UNAME_RELEASE=3
+	test "x${UNAME_RELEASE}" = x && UNAME_RELEASE=3
 	case "`/bin/arch`" in
 	    sun3)
 		echo m68k-sun-sunos${UNAME_RELEASE}
@@ -635,13 +642,13 @@
 		    sc_cpu_version=`/usr/bin/getconf SC_CPU_VERSION 2>/dev/null`
 		    sc_kernel_bits=`/usr/bin/getconf SC_KERNEL_BITS 2>/dev/null`
 		    case "${sc_cpu_version}" in
-		      523) HP_ARCH="hppa1.0" ;; # CPU_PA_RISC1_0
-		      528) HP_ARCH="hppa1.1" ;; # CPU_PA_RISC1_1
+		      523) HP_ARCH=hppa1.0 ;; # CPU_PA_RISC1_0
+		      528) HP_ARCH=hppa1.1 ;; # CPU_PA_RISC1_1
 		      532)                      # CPU_PA_RISC2_0
 			case "${sc_kernel_bits}" in
-			  32) HP_ARCH="hppa2.0n" ;;
-			  64) HP_ARCH="hppa2.0w" ;;
-			  '') HP_ARCH="hppa2.0" ;;   # HP-UX 10.20
+			  32) HP_ARCH=hppa2.0n ;;
+			  64) HP_ARCH=hppa2.0w ;;
+			  '') HP_ARCH=hppa2.0 ;;   # HP-UX 10.20
 			esac ;;
 		    esac
 		fi
@@ -680,11 +687,11 @@
 		    exit (0);
 		}
 EOF
-		    (CCOPTS= $CC_FOR_BUILD -o $dummy $dummy.c 2>/dev/null) && HP_ARCH=`$dummy`
+		    (CCOPTS="" $CC_FOR_BUILD -o $dummy $dummy.c 2>/dev/null) && HP_ARCH=`$dummy`
 		    test -z "$HP_ARCH" && HP_ARCH=hppa
 		fi ;;
 	esac
-	if [ ${HP_ARCH} = "hppa2.0w" ]
+	if [ ${HP_ARCH} = hppa2.0w ]
 	then
 	    eval $set_cc_for_build
 
@@ -697,12 +704,12 @@
 	    # $ CC_FOR_BUILD="cc +DA2.0w" ./config.guess
 	    # => hppa64-hp-hpux11.23
 
-	    if echo __LP64__ | (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) |
+	    if echo __LP64__ | (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) |
 		grep -q __LP64__
 	    then
-		HP_ARCH="hppa2.0w"
+		HP_ARCH=hppa2.0w
 	    else
-		HP_ARCH="hppa64"
+		HP_ARCH=hppa64
 	    fi
 	fi
 	echo ${HP_ARCH}-hp-hpux${HPUX_REV}
@@ -807,14 +814,14 @@
 	echo craynv-cray-unicosmp${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/'
 	exit ;;
     F30[01]:UNIX_System_V:*:* | F700:UNIX_System_V:*:*)
-	FUJITSU_PROC=`uname -m | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz'`
-	FUJITSU_SYS=`uname -p | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/\///'`
+	FUJITSU_PROC=`uname -m | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz`
+	FUJITSU_SYS=`uname -p | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/\///'`
 	FUJITSU_REL=`echo ${UNAME_RELEASE} | sed -e 's/ /_/'`
 	echo "${FUJITSU_PROC}-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}"
 	exit ;;
     5000:UNIX_System_V:4.*:*)
-	FUJITSU_SYS=`uname -p | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/\///'`
-	FUJITSU_REL=`echo ${UNAME_RELEASE} | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/ /_/'`
+	FUJITSU_SYS=`uname -p | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/\///'`
+	FUJITSU_REL=`echo ${UNAME_RELEASE} | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz | sed -e 's/ /_/'`
 	echo "sparc-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}"
 	exit ;;
     i*86:BSD/386:*:* | i*86:BSD/OS:*:* | *:Ascend\ Embedded/OS:*:*)
@@ -896,7 +903,7 @@
 	exit ;;
     *:GNU/*:*:*)
 	# other systems with GNU libc and userland
-	echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr '[A-Z]' '[a-z]'``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-${LIBC}
+	echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr "[:upper:]" "[:lower:]"``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-${LIBC}
 	exit ;;
     i*86:Minix:*:*)
 	echo ${UNAME_MACHINE}-pc-minix
@@ -919,7 +926,7 @@
 	  EV68*) UNAME_MACHINE=alphaev68 ;;
 	esac
 	objdump --private-headers /bin/sh | grep -q ld.so.1
-	if test "$?" = 0 ; then LIBC="gnulibc1" ; fi
+	if test "$?" = 0 ; then LIBC=gnulibc1 ; fi
 	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
     arc:Linux:*:* | arceb:Linux:*:*)
@@ -1272,6 +1279,9 @@
     SX-8R:SUPER-UX:*:*)
 	echo sx8r-nec-superux${UNAME_RELEASE}
 	exit ;;
+    SX-ACE:SUPER-UX:*:*)
+	echo sxace-nec-superux${UNAME_RELEASE}
+	exit ;;
     Power*:Rhapsody:*:*)
 	echo powerpc-apple-rhapsody${UNAME_RELEASE}
 	exit ;;
@@ -1285,9 +1295,9 @@
 	    UNAME_PROCESSOR=powerpc
 	fi
 	if test `echo "$UNAME_RELEASE" | sed -e 's/\..*//'` -le 10 ; then
-	    if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then
+	    if [ "$CC_FOR_BUILD" != no_compiler_found ]; then
 		if (echo '#ifdef __LP64__'; echo IS_64BIT_ARCH; echo '#endif') | \
-		    (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \
+		    (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \
 		    grep IS_64BIT_ARCH >/dev/null
 		then
 		    case $UNAME_PROCESSOR in
@@ -1309,7 +1319,7 @@
 	exit ;;
     *:procnto*:*:* | *:QNX:[0123456789]*:*)
 	UNAME_PROCESSOR=`uname -p`
-	if test "$UNAME_PROCESSOR" = "x86"; then
+	if test "$UNAME_PROCESSOR" = x86; then
 		UNAME_PROCESSOR=i386
 		UNAME_MACHINE=pc
 	fi
@@ -1340,7 +1350,7 @@
 	# "uname -m" is not consistent, so use $cputype instead. 386
 	# is converted to i386 for consistency with other x86
 	# operating systems.
-	if test "$cputype" = "386"; then
+	if test "$cputype" = 386; then
 	    UNAME_MACHINE=i386
 	else
 	    UNAME_MACHINE="$cputype"
@@ -1382,7 +1392,7 @@
 	echo i386-pc-xenix
 	exit ;;
     i*86:skyos:*:*)
-	echo ${UNAME_MACHINE}-pc-skyos`echo ${UNAME_RELEASE}` | sed -e 's/ .*$//'
+	echo ${UNAME_MACHINE}-pc-skyos`echo ${UNAME_RELEASE} | sed -e 's/ .*$//'`
 	exit ;;
     i*86:rdos:*:*)
 	echo ${UNAME_MACHINE}-pc-rdos
@@ -1401,18 +1411,17 @@
 cat >&2 <<EOF
 $0: unable to guess system type
 
-This script, last modified $timestamp, has failed to recognize
-the operating system you are using. It is advised that you
-download the most up to date version of the config scripts from
+This script (version $timestamp), has failed to recognize the
+operating system you are using. If your script is old, overwrite
+config.guess and config.sub with the latest versions from:
 
   http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess
 and
   http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub
 
-If the version you run ($0) is already up to date, please
-send the following data and any information you think might be
-pertinent to <config-patches@gnu.org> in order to provide the needed
-information to handle your system.
+If $0 has already been updated, send the following data and any
+information you think might be pertinent to config-patches@gnu.org to
+provide the necessary information to handle your system.
 
 config.guess timestamp = $timestamp
 
diff -ru -P -x .mirror -x ls_lR -x Old R-release-orig/tools/config.sub R-patched/tools/config.sub
--- R-release-orig/tools/config.sub	2016-01-08 00:15:04.000000000 +0100
+++ R-patched/tools/config.sub	2016-06-21 17:45:14.092455000 +0200
@@ -2,7 +2,7 @@
 # Configuration validation subroutine script.
 #   Copyright 1992-2016 Free Software Foundation, Inc.
 
-timestamp='2016-01-01'
+timestamp='2016-06-15'
 
 # This file is free software; you can redistribute it and/or modify it
 # under the terms of the GNU General Public License as published by
@@ -643,6 +643,14 @@
 		basic_machine=m68k-bull
 		os=-sysv3
 		;;
+	e500v[12])
+		basic_machine=powerpc-unknown
+		os=$os"spe"
+		;;
+	e500v[12]-*)
+		basic_machine=powerpc-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=$os"spe"
+		;;
 	ebmon29k)
 		basic_machine=a29k-amd
 		os=-ebmon
@@ -1382,7 +1390,7 @@
 	      | -nindy* | -vxsim* | -vxworks* | -ebmon* | -hms* | -mvs* \
 	      | -clix* | -riscos* | -uniplus* | -iris* | -rtu* | -xenix* \
 	      | -hiux* | -386bsd* | -knetbsd* | -mirbsd* | -netbsd* \
-	      | -bitrig* | -openbsd* | -solidbsd* \
+	      | -bitrig* | -openbsd* | -solidbsd* | -libertybsd* \
 	      | -ekkobsd* | -kfreebsd* | -freebsd* | -riscix* | -lynxos* \
 	      | -bosx* | -nextstep* | -cxux* | -aout* | -elf* | -oabi* \
 	      | -ptx* | -coff* | -ecoff* | -winnt* | -domain* | -vsta* \
@@ -1399,7 +1407,7 @@
 	      | -morphos* | -superux* | -rtmk* | -rtmk-nova* | -windiss* \
 	      | -powermax* | -dnix* | -nx6 | -nx7 | -sei* | -dragonfly* \
 	      | -skyos* | -haiku* | -rdos* | -toppers* | -drops* | -es* \
-	      | -onefs* | -tirtos*)
+	      | -onefs* | -tirtos* | -phoenix*)
 	# Remember, each alternative MUST END IN *, to match a version number.
 		;;
 	-qnx*)
@@ -1531,6 +1539,8 @@
 		;;
 	-nacl*)
 		;;
+	-ios)
+		;;
 	-none)
 		;;
 	*)

Places

File R-release.diff of Package R (Revision 253a5ee4fa4f3426e7abc1ba63e3d0b0)

Places