[MITgcm-devel] useSETRLSTK

[Jean-Michel Campin]

Hi,

> In terms of model improvements, it would be nice if we removed the size
> declarations from EEPARAMS.h (and put them in a smaller, separate
> EESIZE.h or similar)

I vote "No":
a) EEPARAMS.h does not need to be changed so often (the last time
was 2 years ago, to add useCoupler)
b) the number of files that currently have "include EEPARAMS.h"
and would need to be changed is huge.
c) there is a simple way to avoid the problem that Martin found,
and this is to run testreport before checking in stuff.

Now, I would prefer to have a rather small number of
MAX_NO_PROCS & MAX_NO_THREADS in eesup/inc/EEPARAMS.h
and have a copy of EEPARAMS.h in my customized code directory
when I am about to run with a large number of processors or threads,
rather than the opposite. The only reason why there is
an EEPARAMS.h file for the aim experiments is actually to
set MAX_NO_THREADS to a smaller number than the default.

Jean-Michel