Closed giovanni-rosotti closed 7 years ago
I think this is now ready. It's not passing the tests because it's going overtime with clang, but it's fine with gcc... I'll disable the extra checks and see if we can stay on time. When this is gone, we should think how to reduce the run time on travis because it's painful
I was looking to acccept this, but it seems there is still a problem with MPI...
I swear that when I looked at it this afternoon they were all green ticks! Anyway, I'll give a look at it now...
This is now fixed. I've done some cleanup in the last commit, so provided that travis gives green light, from my side it's ready to go. One of the assertions was too stringent; and then there was a problem with load balancing with the meshless that actually was there since a long time. For the record, it had already been fixed in newparticles2. This is why pull requests should be closed as soon as possible, or we are going to fix every problem twice and waste a lot of time in this way.
I'm happy with this. I'll merge it and close #77
Solves #77