[MITgcm-devel] ad test on lab_sea exp.

[Jean-Michel Campin]

Hi Patrick,

got a lot of recomputation when building mitgcmuv_ad for the lab-sea exp:

- was not the case with Saturday (noon) code (run fine, take ~ same time
 as before)
- but when I take the code from Monday evening:
 co -D "2007-04-23 22:00" -d monday MITgcm
which includes all the KPP modifs + pkg/seaice recent modifications
from the week-end (excluding thoses who affected the forward run
and the output.txt)
got plenty of "TAF RECOMPUTATION" message in taf_ad.log
and the run took for ever to finish (I killed it, but it took ~ 3.h
last night on faulks, instead of ~ 1 minute before).

If someone whants to have a look at this version of the code, 
it's in faulks:/scratch/jmc/monday/verification/lab_sea/build

Jean-Michel