In the spirit of unification, what about putting it in $AMBERHOME -- we'd
have to change all of AmberTools Makefiles, too, but if we're already going
to change all of Amber's...?
In any case, the change is small and Ross's second is encouragement enough
for me to push in the tiny commit (it's easily reversible). I'll work on
the other idea as well and we'll see how it goes.
Thanks,
Jason
On Wed, Feb 16, 2011 at 2:59 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
> I second this!!! Editing the config.h file in two places can be a pain
> anyway. Although rather than a symlink would it not be better to update all
> the AMBER makefiles to point to AMBERTools/src/config.h ? - More work but
> perhaps ultimately cleaner.
>
> All the best
> Ross
>
> > -----Original Message-----
> > From: Jason Swails [mailto:jason.swails.gmail.com]
> > Sent: Wednesday, February 16, 2011 11:35 AM
> > To: AMBER Developers Mailing List
> > Subject: [AMBER-Developers] config.h
> >
> > Hello,
> >
> > Does anyone have a problem with turning the config.h file in
> > $AMBERHOME/src
> > (which is copied at the end of configure) into a symlink (or hard link)
> in
> > the Amber source directory? I can't think of many good reasons to make
> it
> a
> > standalone copy (since I don't think the two files should really differ,
> but
> > others may have different opinions). As far as building Amber, since
> > AmberTools libraries are built for Amber to link to, changing config.h in
> > $AMBERHOME/src and having that *not* fix the issue (because it happened
> > in
> > PBSA or SQM which is using AmberTools' config.h) is confusing and
> annoying
> > (and counterintuitive I think).
> >
> > I'm not sure whether it should be a hard link (you can delete the
> AmberTools
> > config.h and be fine) or a soft link (delete the AmberTools config.h and
> the
> > one in Amber points to nothing). I think that more depends on how
> > standalone Amber is from AmberTools (and how we want to portray that,
> > too).
> >
> > In the case we want to go the way of the link, I have a (2-line) commit
> > ready to push.
> >
> > Thoughts?
> >
> > Thanks!
> > Jason
> >
> > --
> > Jason M. Swails
> > Quantum Theory Project,
> > University of Florida
> > Ph.D. Candidate
> > 352-392-4032
> > _______________________________________________
> > AMBER-Developers mailing list
> > AMBER-Developers.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber-developers
>
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Wed Feb 16 2011 - 12:30:06 PST
