Hi Peter, I want to run the system tests on a number of different architectures/CUDA installations to make sure we have stability. Looking at the documentation (http://docs.openmm.org/latest/userguide/library.html#system-tests) I downloaded the tests but I have a few questions/concerns. Firstly I'm ...

I think the way to do this is to build the dists individually by running:
and/or
and then copying those to your python installation or put on your python path.

Thanks Peter,
I guess I will just plan on building myself if I need to circumvent conda.

For the new conda installation process, is there going to be a way to specify to install openmm built with a certain version of CUDA without actually having to install 'cudatoolkit' (and having the dependency solver figure it out presumably) through conda? I liked how the labels in the old omnia rep...

Thanks for the clarification and tip that helps.

Is it possible to have a context without periodic boundary conditions? That is to have a value that explicitly indicates that there are no boundaries at all. For example I make an empty system and the default periodic boundary conditions are an arbitrary box: In [1]: import simtk.openmm as omm In [3...

First of all, thanks for the improvements in this release they will be very important to applications like weighted ensemble where we have to start a lot of new simulations and build up these objects from scratch. In the release notes you mention that creating new contexts in some special cases can ...

Much appreciated!

Do you have packages built against multiple CUDA versions for OpenMM 7.2?

Right which is why I'm asking on the forum to see if anyone had any ideas on making it work. There is no NVIDIA package for Ubuntu 18.04 yet, and I read elsewhere that there are conflicts and thus can't be made to work.