VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Fri Jun 27 2008 - 13:54:36 CDT

On Fri, 27 Jun 2008, Marylou Kunkle wrote:

marylou,

MK> I am trying to compile VMD from source on Mac OS X 10.5.
MK>
MK> I am kind of new to Mac in general - just for a warning. However, I have

ok. i don't use/own a mac. for the most part, however, OS-X is not
much different from a unix system, apart from the default file system
being case-preserving, but case-insensitive...

MK> read all the documentation a million times over, but I can't seem to get
MK> these libraries to compile or ... perhaps they're all in the wrong place.
MK> Either way, it yells at me that the plugins are not compiled and copied to
MK> 'distrib'. The plugins won't compile because of a number of things, I'm

you _have_ to compile the plugins. particularly the molfile_plugin
library. it is essential for VMD to work.

MK> sure. I keep getting the error about it being unable to find -lnetcdf ... I
MK> don't know how to fix it. I have unsuccessfully tried to set the PLUGINDIR

you have to compile/install the netcdf package to compile the amber
plugin. if you know your way around makefiles, you can comment
that one out for the time being and remove the references to the
netcdf library.

MK> env. variable. Or I have set it, and that's not what is wrong at all .. I am
MK> not sure.

MK> I started by downloading and installing gcc and a whole bunch of libraries:
MK> fltk, opengl, tcl, and now i downloaded netcdf but i don't know what to do
MK> with the files it gave me. Then I tried a number of crazy things to try to
MK> get things to compile, but truthfully, I don't know what I'm doing.

it is hard to give you any help with that. building a large
package software like VMD, can be a tricky issue. trying this
with not much experience and on a platform that you don't know
is a task worthy of a don quixote, i.e. truely herioic but
with little impact.

MK> Why am I trying to compile from source on a Mac? Because we added some stuff
MK> to VMD and I am trying to port our version of this program to Mac and
MK> Windows. Just in case there's any confusion as to why I am putting myself
MK> through all this. :)

have you thought about contacting VMD developers to simply integrate
your software directly into the regular VMD release? depending on
how close the code is tied to the current development platform
(which one, btw?) it may be a minimal effort and the porting building
effort would be offloaded to people with experience on cross-platform
builds.
ok, john is currently unavailble for roughly another couple of weeks.
but if you are willing to deal with a "sancho panza" instead, i can
give it a look and give some comments based on my experiences in
programming and building VMD.
 
MK> Can anyone give me a nice step-by-step tutorial of where to start? I guess
MK> what is on the website really isn't detailed enough for me.

hmmm...., the available information is IMO quite sufficient to
describe the process. if you have trouble it means that you are
lacking a more fundamental understanding of the whole process of
compiling, installing and packaging process of software. this can
hardly be the withing the scope of VMD specific documentation.
moreover, it is very system specific and dependent on local practices.
e.g., my build system is a different from the one that john uses,
since i'm using "native" libraries and don't compile for many platforms
at a time. finally, it is actually quite difficult for a person
that has little problems with the build of VMD to imaging what
kind of problems your are specifically facing, so writing that
kind of documentation would be very difficult (in fact you would
probably be most qualified after you found out what you are
currently missing).

if you are not doing this already, i also recommend to build
from the cvs sources as that reduces the effort to port to the
next release version and eases the process of communicating
with the developers.

cheers,
   axel.

MK>
MK> Thank you,
MK>
MK> Marylou
MK> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.