Recent changes in OpenMC

Hi all,

Since the version 0.8.0 release of OpenMC in late July, there have been a number of substantial changes in the code that I wanted users and developers to be aware of. I’ve started putting together a wiki with a list of these changes. Perhaps the biggest change is that we are now using an HDF5 format for nuclear data rather than ACE files, so to use the current developmental version of OpenMC (and future releases), you will need to convert your ACE files. This can be done with the Python API within OpenMC or via a number of utility scripts. Both the Python API and HDF5 data format have been somewhat unstable; I know this may cause pain for users, especially those who are living on the cutting edge and using the develop branch as opposed to a release version. However, we’re trying to make these improvements now so as to reach a point of stability sooner rather than later.

Regarding HDF5 data – if you are using the develop branch and need a quick source of working data, have a look at the .travis.yml file in the root directory of the repository which has a URL to a tar file with data that is used by our automated testing.

I also wanted to pass along recommendations on installing OpenMC if you are using conda for Python since I know some of you have had troubles with it recently:

  • Do not use the OpenMC binary from conda-forge as it is currently broken

  • Download the source, either from a release tarball or through git

  • When configuring use:

  • cmake -DCMAKE_INSTALL_PREFIX=${CONDA_PREFIX} …- If you’re using version 0.8.0, install h5py=2.5 from conda. Make sure you are referring to the stable documentation.

  • If you’re using the develop branch, use any version of h5py. Make sure you are referring to the latest documentation.
    As always, if you have questions, please post a message to the user’s or developer’s mailing list and we’ll try to help as best we can.

Best regards,
Paul