We could always distribute a supplementary test suite as a separate
download, since they do provide good examples, and just have a barebones
set of tests included in the tarballs to, as Scott put it, make sure stuff
works. (We could also only include the "typical" tests, excluding other
types of simulations, like constant pH, PIMD, REMD, etc.).
Just a thought. I won't even rank that idea in terms of how good I think
it is :).
Jason
On Wed, Dec 21, 2011 at 5:10 PM, Scott Brozell <sbrozell.rci.rutgers.edu>wrote:
> Hi,
>
> On Wed, Dec 21, 2011 at 01:47:43PM -0500, David A Case wrote:
> > I wonder if we should
> > consider having a test suite for developers, but perhaps a smaller one
> > (leaving out a few really big tests) for users. Or, are there existing
> > tests that are actually redundant for everyone?
> >
> > Among other problems, we have
> > users with limited bandwidth Internet connections.
> >
> > I think the first step is for people to look for things (especially but
> not
> > exclusively in the test directories) that might be removed. You can edit
>
> Yes to all of the above which is basically spring cleaning.
>
> As far as the test suite, my feeling is that we should distribute
> one focused to answer:
> 1. did program x get installed,
> 2. does program x work for tried and true basic functionality,
> 3. does program x work with the latest and greatest new functionality.
>
> The first two could amount to one or a very small number
> of tests. The third should probably be close to the
> developer test suite. And the dev suite should be
> downloadable and fitable to be runable.
>
> A good reason not to do this would be empirical data
> showing that a myriad of tests found a myriad of problems post release.
> My feeling is that the reflector does provide such data.
> Another good reason not to do this would be if the tests
> really do provide good examples.
>
> scott
>
>
> _______________________________________________
> 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 Dec 21 2011 - 14:30:04 PST
