1. em: Completely on board with using algebra::Grid3D as a base. Will
simplify many things. Lets talk about it offline.
2. I think a *very* important feature for 1.1 is many many examples/
tutorials and applications (systems). It may very well be the show
stopper for the release :)
We should decide on the systems for the 1.1, that are also
aligned with the IMP paper.
Keren.
On Nov 5, 2010, at 11:08 AM, Daniel Russel wrote:
Given that it has been a while since IMP 1.0 was released and IMP
has improved a lot, I think we should have an IMP 1.1 release in the
near future (say, on the order of a month). There are a few modules
with fairly active development, which can either stabilized for a
release or skipped from the release. The status of each modules as I
understand it (ok means "no implementation plans/known important
bugs"):
- build system: ok
- kernel: ok
- cgal: volume/surface are computations are not heavily tested, but
they are not a show stopper
- algebra: ok
- bullet: ok
- display: ok
- em2d: active development, Javi, what are your plans/goals?
- modeller: ok
- restrainer: ok
- domino: should probably go away for the release
- example: ok
- multifit: active development. Keren, what are your plans/goals?
- saxs: ok
- atom: ok (recent development in the protein-ligand score, more
bugs might be discovered there)
- container: ok
- domino2: ok (poor handling of nbls with rigid bodies, but I don't
think it is a show stopper)
- gsl: ok
- statistics: ok
- benchmark: ok
- core: ok
- em: discussed plans for moving to algebra::Grid3D and moving
algorithms to algebra where they can be more widely used, but no
implementation. This would be nice, and should be relatively quick,
but not a show stopper, I think.
- misc: ok
- openmm: should not be part of release
I'll make a pass through before the release to add more comments to
example code (others feel free to contribute, please, please
please :-)