I've been working on updating the merge between yt-4.0 and master.
As it stands:
* All travis tests and appveyor tests pass. This includes the "minimal answers," which I personally verified were correctly updated. (You can see the HTML outputs of the diffs here: http://use.yt/upload/3e427ec7 ) * py2 tests are all failing, which is to be expected. * py3 answer tests are failing, which is *also* to be expected.
This means that the answer tests, which necessarily should fail because *we have changed behavior for SPH/n-body results*, are the remaining item to verify. I am working to prepare a set of verification tests for the results which I can manually check off, which I will put in the pull request as I check them off with manual verification.