[ghemical-devel] mopac7-1.14rc1 released

Daniel Leidert daniel.leidert.spam at gmx.net
Tue Aug 5 10:58:23 EDT 2008


Am Dienstag, den 05.08.2008, 13:36 +0300 schrieb Tommi Hassinen:
> Should be OK now. As mentioned in the INSTALL file, the configuration should now be made like this:
> 
>         export F77='g77 -fno-automatic'
>         ./configure
>         make
>         make install
> 
> and also this is OK:   export F77='gfortran -std=legacy -fno-automatic'
> 
> If the configure-script can be made to detect presence of g77 or
> gfortran automatically it would be nice ;

That's pretty easy. I attached a patch also fixing the configure.ac [1],
the order in autogen.sh and removing some useless stuff - BTW: AR is
internally defined by GNU make - is there a reason to determine it by
configure? IIRC dito for SH)

>  however I can't do such improvements myself so I wait for patches
> and/or bugreports and then just drop "rc1" from filename if nothing
> new appears.

Hm. gfortran refuses to compile symtrz.f. If I apply
http://svn.debian.org/wsvn/debichem/unstable/mopac7/debian/patches/03_fix_FORTRAN_source.dpatch?op=file&rev=0&sc=0 and revert your changes and then comment the line:

NAMES=IFRA

then it compiles. However, I don't know, what is the correct solution
for the subroutine R00010. There is an assignment of character and
integer values. Maybe the character-assignment (see above) is wrong or
the integer assignment. Do you know more about the routine?

Maybe NAME should be assigned with IFRA and NAMES is an integer? Could
this be? I simply did this in my patch. Please test.

[1] VERSION is defined by AC_INIT. Your AM_INIT_AUTOMAKE and AC_OUTPUT
usage is deprecated for a while now.

Regards, Daniel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mopac7.patch
Type: text/x-patch
Size: 5230 bytes
Desc: not available
Url : http://www.bioinformatics.org/pipermail/ghemical-devel/attachments/20080805/cb0cdb6c/attachment.bin 


More information about the ghemical-devel mailing list