[MITgcm-devel] [MITgcm-cvs] MITgcm/pkg/seaice CVS Commit

[Gael Forget]

Hi Dimitris,
no worries. I am guilty myself of not paying close 
enough attention to the beginning of the discussion.
I think I am all but ready to modifying the lab_sea 
and offline_exf_seaice adjoints. I will let tonight's
testing complete first. I am not sure if you still have
pending checkins. Could you give me a heads up
when you are done? I will then do my checkin,
then we would do a tag, and I will remove the 
legacy code itself next week.
Cheers,
Gael

On Dec 21, 2012, at 1:19 PM, Dimitris Menemenlis wrote:

> OK.  I will proceed with warning then.
> Sorry my changes had so many unintended consequences.
> 
> Dimitris Menemenlis
> 
> On 12/21/2012 10:12 AM, Gael Forget wrote:
>>> I keep my preference (warning instead of error+stop),
>>> >adding that the default SEAICE_SIZE.h allows to use
>>> >several (up to 7) ice-categories. So, it'not that
>>> >the code is gone (like SEAICE_OLD_AND_BAD_DISCRETIZATION).
>>> >I just have the impression that many users have some difficulties
>>> >in updating their seaice params+options setting, and
>>> >don't think this stop helps (as long as there is a clear warning),
>>> >but it's only my impression.
>> Makes sense.
>> Gael
>> 
>>> >
>>> >Cheers,
>>> >Jean-Michel
>>> >
>>> >On Fri, Dec 21, 2012 at 12:05:31PM -0500, Gael Forget wrote:
>>>> >>Hi dimitris,
>>>> >>that CPP is only retired since you retired it. It was setting the
>>>> >>SEAICE_multDim default, which was all that was needed
>>>> >>to preserve the previous behavior of pkg/seaice. Since you
>>>> >>effectively changed the default of SEAICE_multDim, any user
>>>> >>of SEAICE_MULTICATEGORY will have to change his namelist,
>>>> >>otherwise his results will change. Just like what Jean Michel
>>>> >>did to fix the verification experiment. Except those users may
>>>> >>not notice the warning, and thus end up with different results.
>>>> >>So a stop may be needed, to avoid tricking users. This is
>>>> >>what I was trying to avoid with the code you removed.
>>>> >>Cheers,
>>>> >>Gael
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel