compiling 0.62.3 for SunOS 4.1.4 and Solaris 2.6

Ben Bolker ben@eno.Princeton.EDU
Tue, 15 Sep 1998 15:46:24 -0400 (EDT)

  Just a quick report on compiling R 0.62.3 for SunOS 4.1.4 and Solaris
2.6 (both with gcc 2.7.2.something), in hopes that it'll be useful to


     - I followed Charles Berry's forward of Ross Ihaka's suggestion about
putting fake #defines for gamma and lgamma in src/include/Mathlib.h (this
works best on the second pass, after the graphics library (which also
contains a "gamma" function) has compiled). 

     - I got regex-0.12 from, made
it into libregex.a, hacked HAVE_REGCOMP in src/include/Platform.h, and
added /usr/local/lib/libregex.a to RLIBS in src/main/Makefile (there
didn't seem to be a way to get configure to recognize it automatically). 

     - got readline-2.2 and installed it.

  I have to say I have mixed feelings about readline and regex.  It
certainly makes sense to use these existing resources, but it does make R
harder to install "out of the box" on systems (like SunOS) that don't have
them by default ... (I know, if SunOS had a decent packaging system we
could put them into package dependencies -- oh well.) 


  I don't know if it's just my system (which has been administered by
various moderately clueless people in the past), but I had a number of
troubles with gcc and Solaris's default include files.  Several times
important definitions were located in chunks of the include files that
didn't get included by default. 

   - I added "typedef int sigjmp_buf[_SIGJBLEN];" to src/include/Defn.h by
hand; had to make sure that the gcc version of math.h got included rather
than the default Solaris file; and ran the compilation of the stuff in
src/main with -D__EXTENSIONS__ added to CFLAGS in Makeconf.  I *think*
that adding -D__EXTENSIONS__, or alternately setting -D__STDC__=0, to the
CFLAGS line in Makeconf, might have fixed all of my problems at once. 

Benjamin Bolker          Dep't of Ecology and Evolutionary Biology,    Princeton University
tel: (609) 258-6886,     fax: (609) 258-1334

r-devel mailing list -- Read
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: