A few changes to avoid compilation errors with the PGI compiler. This was mainly caused by variables with limited scope and min() with 3+ arguments.
Change output of number of cells to a long long for very large simulations with >2 billion cells on a level.
Apply Godunov's formula to all hydro solvers instead of just PPM.
Breaks several test problems in the test suite, beyond the two we expect to fail (SphericalInfall and DrivenTurbulence3D). It breaks the CoolingTest (three diff. variants), ExtremeAdvectionTest, and PhotonTest.
I don't believe so. If I look at the comments in the hudson run 256 (http://srs.slac.stanford.edu/hudson/job/enzo-dev/256/parameters/), it says "will test PR 285 by John Wise", which is this PR. The hash (ENZO_REV=a5a44797a8b6) also matches the last commit for this PR. The link you've included points toward PR #284. Why the numbers are out of order is anybody's guess, but you should verify that what I'm saying makes sense - otherwise basically all of the messages I posted this morning are suspect.
I was looking at the commit descriptions in (http://srs.slac.stanford.edu/hudson/job/enzo-dev/257/) which are from PR 285. But the changeset hash and comment in the parameters link you posted are from PR 284. I think something is going wrong with hudson to mix up these two PRs.