[MITgcm-devel] Re: [MITgcm-cvs] MITgcm/verification/fizhi-gridalt-hs/input CVS Commit

[Ed Hill]

On Thu, 2004-11-04 at 22:15, Andrea Molod wrote:
> hi ed,
> 
> > 
> > OK, I understand your point.
> > 
> > So could we please change the code so that the "datetime" input file
> > isn't over-written every time that the code runs?  The problems are: 
> > 
> >   1) breaks the automated testing scripts for every run except 
> >      the first one and
> > 
> >   2) *none* of the other examples have this sort of behavior.
> > 
> > So could we have it write a "datetime_new" file instead?
> > 
> 
> i do not want to change the file name - the exact point is that it 
> have the same name for every run - it acts as a sort of pickup for
> the date. and there is a reason why none of the other tests need
> to do this - fizhi needs to know the 'real' simulated date and time,
> which may or may not be computed properly from the 'start time' of
> the run. so this little pickup is needed until everything gets done
> in mnc and the date is kept as metadata there. all of the fizhi tests 
> will work like this.
> 
> can't the test run make sure that is has the file that is checked
> in to the repository? this issue is equivalent to making sure that
> the 'nIter0' parameter in the data file is correct.....
> 
> if this is not possible (making sure that 'datetime' is the one in
> CVS), then we should chat with jean michel and chris about how to
> set up the test so that it works.

Hi Andrea,

This "datetime" is a pickup file like any of the other pickups.  So it
seems like a poor idea to over-write it.  The overwrite approach is
completely inconsistent with the way *all* of the other MITgcm pickups
are performed.  And its mighty inconvenient--it forces the user to do a
fresh check-out or to keep extra copies of input files around if they
ever intend to run that verification more than once.

In my opinion, we should be consistent with all the other verification
tests and edit "fizhi_write_datetime.F" so that it has:

      open(10,file='datetime_next',form='formatted')

or some other name that doesn't overwrite the input file.

And yes, I agree that this should be reviewed by the group so I've
forwarded it to the devel list.

Ed

-- 
Edward H. Hill III, PhD
office:  MIT Dept. of EAPS;  Rm 54-1424;  77 Massachusetts Ave.
             Cambridge, MA 02139-4307
emails:  eh3 at mit.edu                ed at eh3.com
URLs:    http://web.mit.edu/eh3/    http://eh3.com/
phone:   617-253-0098
fax:     617-253-4464