[MITgcm-devel] ptracers seem to be broken for more than one tile

[Martin Losch]

Hi there,

contrary my prior suspicion there seems to be a problem with ptracers. 
I wonder when someone has looked at this for the last time, but the 
global_ocean.90x40x15 experiment, which seems to be the only 
(forward-)test for ptracers (except for dic_example), produces a 
results that is only the same as salinity for half of the domain.
Also the new testreport fails for dic_example on more than one 
processors on the ptracers-part, another indication that there's 
something wrong with ptracers (or with gchem_monitor, but I doubt 
that).
Has anyone a good idea as to where this package could have gotten 
broken (to make life easier for me).

Martin