@craigappl, I was the mentor on this GSoC project. If I remember correctly:
- the code we merged works (in theory), and can fetch patients from the MPI, and update them there
- it was never properly tested, because we could never get OpenEMPI working right in a test setup
- I don’t know of anyone who has ever used it in production
I have only limited secondhand experience with OpenEMPI, and no experience with other MPIs, but my experience with OpenEMPI has been negative. If you have a choice between different MPI tools, the fact that some work was done on OpenMRS-OpenEMPI integration should not be what drives your decision. (This code is built to be able to swap out OpenEMPI with a different MPI back-end.)