If nab doesn't use AMBERHOME, then I believe that the location is hardcoded
and requires you to recompile.
Moreover, although nab and related tools look like they only have C code and
a C compiler should be good enough, there are 2 issues. First is that the
configure script will fail without a valid fortran compiler. Therefore, if
you can install a Fortran compiler (even g77 and make a gfortran alias),
you'd be much better off. Second, nab calls blas routines -- the blas
routines are written in Fortran77, so you need to be able to compile that
code in order to build nab. You can also type "make nabonly" to build just
nab. Note the g77-linking trick will *not* work with the rest of Amber and
AmberTools that uses Fortran 90 (i.e. PBSA).
To get Amber to compile with nonstandard compilers, you'll have to play
around with the configure file (or create a config.h file by hand).
Hope this helps,
Jason
On Sat, Feb 12, 2011 at 5:47 PM, Anselm Horn <
Anselm.Horn.biochem.uni-erlangen.de> wrote:
> Dear Prof. Case,
>
> thanks for your quick response!
>
> > Nab is a compiler, not a stand-alone prgram. It converts your input
> program to C, then relies on your system compiler to convert this to machine
> code. Bottom line: if you don't have compilers on some machine, nab won't
> do you any good.
>
> I apologize for my wrong phrasing: On the machine, where I intend to run
> nab, I do not have the complete compiler suite (Fortran + C) to build
> Amber11 and the AmberTools1.4 from scratch, but I do have the standard
> gcc compiler installed, so that should be fine.
> However, invoking nab (residing in /home/horn/amber11/...) yields the
> error:
> /apps/soft/amber11/bin/ucpp: No such file or directory
>
> This path was the directory structure from the machine where AmberTools
> and Amber were compiled originally.
>
> So I wonder whether it is possible to tell nab where to find its ucpp
> preprocessor in a different environment. I tried some command line flags
> of nab (-B , -std, -specs, --sysroot) but did not succeed.
>
> Can you help me there?
>
> Many thanks in advance!
>
> Regards,
>
> Anselm
>
> Bioinformatik
> Friedrich-Alexander-Universität Erlangen-Nürnberg
> Germany
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Feb 12 2011 - 22:00:02 PST
