[MITgcm-devel] Problems with diagnostics package and mnc

[Jean-Michel Campin]

Hi,

myTime & myIter passed as arguments was working until now.
I also though that passing thing as argument was 
"the way to go". So may-be we can continue like this for 
myTime & myIter ?

Can you give example where this does not work (apart from internal
mnc S/R) and needs to be changed ?

The other thing is that we need to know what time it is
when a real calendar is used (cal or with fizhi).
I would prefer to have Patrick & Andrea opinion on this 
point before making changes in the way the time is kept.
Here my concern is the risk to end-up with 4 or 5 different ways 
to keep a record of the time, with lot of overlap,
that might not be very clear at the end.

Jean-Michel

> On Wed, 2004-12-29 at 09:14 -0500, Chris Hill wrote:
> > Hi Ed,
> >
> >   There is a variable that is passed around as an argument.
> >   Its generally called myTime or myCurrentTime. It holds the simulation
> > time in seconds. You should use that for the basic time axis in netCDF
> > output.
> 
> Hi Chris & Jean-Michel,
> 
> Could we (please?) put it within a common block?
>  
> How about model/inc/PARAMS.h?
> 
> The reason I ask is that it could (or should?) be a NetCDF coordinate
> variable and it would make it a lot easier for me not to have to pass a
> variable down through so many function calls (eg. see the call graph for
> pkg/mnc/mnc_cw_cvars.F)
> 
> Ed