Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!cs.utexas.edu!uunet!sco!seanf From: seanf@sco.COM (Sean Fagan) Newsgroups: comp.sources.games.bugs Subject: Re: GCC on SCO UNIX was Re: Nethack on SCO UNIX Message-ID: <7962@scolex.sco.COM> Date: 1 Oct 90 22:22:06 GMT References: <1090001@hppcgelo.hp.com> <6370001@hppcmart.HP.COM> <7938@scolex.sco.COM> <89@unigold.UUCP> Sender: news@sco.COM Reply-To: seanf (Sean Fagan) Distribution: usa Organization: The Santa Cruz Operation, Inc. Lines: 29 [Hmm. 7938@scolex.sco.com was probably me, but I don't remember it 8-).] In article <89@unigold.UUCP> kianusch@unigold.UUCP (Kianusch Sayah Karadji) writes: >It didn't like the archive it producd for gnulib... >I took this part out... but it did't help... The problem is that the makefile will try to run ranlib on the COFF objects. Not necessary, and, in fact, ranlib doesn't like it. Add a '-' to the front of all of the lines that run ranlib; this will cause it to ignore the errors (ranlib won't modify the file in any way, so the error is ignorable). >I think it didn't like the compiler ... >I'm using (it's the one comming with SCO's develpement kit) I've had no problems getting gcc working under the SCO devsys. Sometimes, it takes a bit of work, though, because of the long macros gcc likes to use. Generally, most of the problems in that respect were solvable by using rcc on the offending module, or running the module through a preprocessor (gcc-cpp, /lib/cpp, or cc -E, depending), and then compiling the output. If you use rcc, be sure to add '-DM_UNIX -DM_COFF' to the command line; I would suggest you put this into the things gcc-cpp automatically defines. -- -----------------+ Sean Eric Fagan | "Never knock on Death's door: ring the bell and seanf@sco.COM | run away! Death really hates that!" uunet!sco!seanf | -- Dr. Mike Stratford (Matt Frewer, "Doctor, Doctor") (408) 458-1422 | Any opinions expressed are my own, not my employers'.