From chemistry-request@ccl.net Mon Dec 2 10:25:27 1991 Date: Mon, 2 Dec 91 10:00:20 -0500 From: nobody@Kodak.COM To: "chemistry@ccl.net"@Kodak.COM Subject: Re molecular modeling report. >From: NAME: Adi M. Treasurywala FUNC: Biophys. & Compu. Chem. TEL: (518)445-7042 To: "chemistry@ccl.net"@KODAKR@MRGATE@WPC Doug, The report that (I think) you were asking about is called Conflicting Trends in Computational Chemistry. It was put out by the Aberdeen Group Inc. 92 State Street, Boston Mass 02109. (617-723-7890). The contact person was (in May 1989!!) Charles T Casale. I did not get a copy but took a look at one at a friends place of work. It seemed to be very much slanted towards Polygen and the IBM solution (both were somehow involved in its sponsorship I later found out.) But it had some very well thought out points too. There is a new update on this report out this year I heard but I have not seen it yet. Perhaps if you get a copy you would like to share your thoughts with the net. Hope this helps, Adi T. (adit@kodak.com) From jkl@ccl.net Mon Dec 2 12:29:35 1991 Date: Mon, 2 Dec 91 11:43:34 -0500 From: jkl@ccl.net Subject: Re: COmpiling individual link of G90 on IBM RS/6000 To: chemistry@ccl.net I am forwarding to the list(jkl@ccl.net) ---------- Begin Forwarded Message ---------- >From bbesler@ouchem.chem.oakland.edu Mon Dec 2 10:45:38 1991 Date: Mon, 2 Dec 91 10:47:50 -0500 From: bbesler@ouchem.chem.oakland.edu (Brent H. Besler) To: chemistry-request@ccl.net Subject: COmpiling individual link of G90 on IBM RS/6000 Status: R The bldg90 script that comes with the RS 6000 version of Gaussian 90 seems to be only capable of compiling the entire program, not individual overlays(say l101, l302, etc.). How do you compile and link individual overlays? ----------- End Forwarded Message ----------- From chemistry-request@ccl.net Mon Dec 2 12:38:13 1991 Date: Mon, 2 Dec 91 12:20:38 EST From: m10!frisch@uunet.UU.NET (Michael Frisch) Subject: Re: COmpiling individual link of G90 on IBM RS/6000 To: chemistry@ccl.net I am forwarding to the list(jkl@ccl.net) ---------- Begin Forwarded Message ---------- >From bbesler@ouchem.chem.oakland.edu Mon Dec 2 10:45:38 1991 Date: Mon, 2 Dec 91 10:47:50 -0500 From: bbesler@ouchem.chem.oakland.edu (Brent H. Besler) To: chemistry-request@ccl.net Subject: COmpiling individual link of G90 on IBM RS/6000 Status: R The bldg90 script that comes with the RS 6000 version of Gaussian 90 seems to be only capable of compiling the entire program, not individual overlays(say l101, l302, etc.). How do you compile and link individual overlays? ----------- End Forwarded Message ----------- 1. You can use the file g90/bsd/g90.make to build individual links: cd $g90root/g90 make -f bsd/g90.make l302.exe will recompile and relink l302, assuming that l302.F is more recent than l302.exe. 2. If you are debugging your own link in your own directory, you can start from the sample makefile given below. It ensures that the compiler flags (e.g., -qextname) are applied consistently: -------------- Beginning of Makefile --------------------- IBM_FC = xlf -lblas MFL_FC = fc TIT_FC = f77 SUN_FC = fc # set FC to be one of the above variables depending on whether you # are on an RS/6000, Multiflow, Stardent Titan, or Sun FC = $(IBM_FC) .SUFFIXES: .SUFFIXES: .o .F # change the following to specify wherever the standard Gaussian # files are kept. GDIR = /g90 .F.o: make -f $(GDIR)/bsd/g90.make $*.o # This should be a list of the links to be built. all: l1102.exe # Each link should have a list of contributing files like this. # Only routines which have been modified need be listed. OBJ1102 = ml1102.o # Each link to be made needs an explicit rule similar to the # one below. If the link is entirely new, the reference to # the standard version of its library should be omitted. l1102.exe: $(OBJ1102) $(FC) -o l1102.exe $(OBJ1102) $(GDIR)/l1102.a \ $(GDIR)/util.a -------------- End of Makefile --------------------- -------