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 :-)
Other issues?
Thanks.
Hi - 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 :-) > > Other issues? > > Thanks. > _______________________________________________ > IMP-dev mailing list > IMP-dev@salilab.org > https://salilab.org/mailman/listinfo/imp-dev
Everything that you guys think is OK, I am on board with, including renaming domino as indicated in other emails.
I very strongly agree with Keren about tutorials and examples. I think that the new Sali Lab members and visitors should be closely consulted on these, since they're in the process of learning everything and we can see if what's there makes sense. Maybe we can have a roundtable next week. Sadly, I might be getting to the point where IMP is intuitive and therefore can't say if it makes sense from a newbie perspective.
Dave
On 11/5/10 12:17 PM, Keren Lasker wrote: > Hi - > 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 :-) >> >> Other issues? >> >> Thanks. >> _______________________________________________ >> IMP-dev mailing list >> IMP-dev@salilab.org >> https://salilab.org/mailman/listinfo/imp-dev >> > _______________________________________________ > IMP-dev mailing list > IMP-dev@salilab.org > https://salilab.org/mailman/listinfo/imp-dev >
On 11/5/10 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
Agreed. We will need some kind of working version of multifit in that release, given the extent to which we have demonstrated the module and promised its existence "soon". Keren, what kind of timescale are we talking about here?
Ben
Given the licensing issues will be resolved, it is mostly done.
On Tue, 23 Nov 2010, Ben Webb wrote:
> On 11/5/10 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 > > Agreed. We will need some kind of working version of multifit in that > release, given the extent to which we have demonstrated the module and > promised its existence "soon". Keren, what kind of timescale are we > talking about here? > > Ben > -- > ben@salilab.org http://salilab.org/~ben/ > "It is a capital mistake to theorize before one has data." > - Sir Arthur Conan Doyle > _______________________________________________ > IMP-dev mailing list > IMP-dev@salilab.org > https://salilab.org/mailman/listinfo/imp-dev >
-
Ben Webb -
Daniel Russel -
Dave Barkan -
Keren Lasker