[MITgcm-devel] bottom_ctrl_5x5 broken

[Jean-Michel Campin]

Hi Patrick,

Did you notice that the bottom_ctrl_5x5 experiment is 
giving NaN for couple a months now.
it was working with checkpoint59d (June 20) but was already broken
at the end of June (tested July 1rst, 6.am).

The unfortunate thing is that testreport was giving the same 
matching number: 16 , 6  no matter of how many NaN were produced
(this testreport problem should be fixed by now, well, will see).

Jean-Michel