[Bug 497622] Review Request: apbs - adaptive poisson boltzmann solver

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


https://bugzilla.redhat.com/show_bug.cgi?id=497622





--- Comment #27 from Tim Fenn <fenn at stanford.edu>  2009-07-23 21:42:23 EDT ---
(In reply to comment #26)
> apbs has support for:
>   --with-mpich=PATH       toplevel MPICH directory
>   --with-mpich2=PATH      toplevel MPICH2 directory
>   --with-lam=PATH         toplevel LAM-MPI directory
>   --with-openmpi          enable OpenMPI compilation
> so I suggest you add these to the package in the future.
> 
> I have suggested an MPI packaging draft and an environment modules packaging
> draft, which would standardize the way MPI stuff is packaged.
> 
> - Add BR: arpack-devel and --with-arpack to enable support for ARPACK.
> 
> - Add BR: python-devel and --with-python to enable support for Python.
> 

Adding --with-arpack enables building of a "driver" binary that is part of the
tools folder, which is just copied to the share directory in the
install-data-local macro, so its not required for APBS.  Would it be best to
set up a -tools subpackage with all the extras? (see here:
http://cardon.wustl.edu/MediaWiki/index.php/An_overview_of_the_APBS_package#Other_tools_distributed_with_APBS
for more info) And I'm also worried about having a binary just called "driver"
lying around - could that be an issue?

Adding --enable-python doesn't do anything other than build ZSI (which it
shouldn't do) - which I realized, python-ZSI should be a requires so
ApbsClient.py runs properly.

I'm working on the remaining issues, but I wanted to check with this before I
finish it off.

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.