[Mitgcm-support] Re: mitgcm random survey!!!

[mitgcm-support at dev.mitgcm.org]

Hi Chris,

Ok, here are some specific, but in no way organized comments. Hope that
you will not take it as an evaluation - there is only  nagging, because I
think, this is what you wanted.
The comments may not apply for the official releases.

- The use of very many subroutine levels makes hard to find out what is
going on, some of them are really unneccessary (i.e. cost_hyd calls only
cost_Theta and cost_Salt, active_write_xy calls only active_write_rl (ecco
only)). Since most of the subroutines are only used once they seemed to be
introduced to enhance the elegance of the
code and could be combined a bit. I find myself giving up to climb further
down. A subroutine that is only used once should consist of more than two
lines of code. I like that filenames correspond to the subroutinenames.

- Use of capitalized letters (i.e lAdInit) is not very  helpful, it
is not strictly handled and TAMC does not  respect it (well ... it's not
a problem)

- the use of switches (#ifdef's) is difficult, because one could easily
make nonsens combinations. At the moment, the system is in some sort of
intermediate state, you receive warnings but you could not rely on
being warned (i.e switching KPP on, gives a warning that implicitdiffusion
is necessary, but you are able to run the Model without
implicitviscosity. In this case, it would be helpful if the switch KPP
changes the default values instead of a warning, so that (I don't know if
this in possible for any case) specifying nothing gives a reasonable
combination.

- Could the  external forcing package be part of the official mitgcmuv?
 Otherwise it is difficult to upgrade to a new release. Ralf's ecco 
upgrade to c30 now includes routines which are called and which calculate
something (lambdas for relaxation), but these values are never used. 

-I had/have problems to run the model on certain machines (i.e. never got
it run on a Sun (does not apply for the verification experiments))

I think that's it for now
Armin

 _________________________________________________________________
Armin Koehl
Physical Oceanography Research Group       phone: +1-858-822-4325
Scripps Institution of Oceanography        fax  : +1-858-534-9820
8605 La Jolla Shores Dr.                   email: akoehl at ucsd.edu
La Jolla, CA 92093-0230, U.S.A.