[MITgcm-devel] testing status

[Jean-Michel Campin]

Hi,

Regarding lab_sea adm results, it turns out that with the new TAF version
(2.9.2) the output is identical to the old one (before June 19, TAF version
2.8.16). I will put back the old output_adm.txt.

Cheers,
Jean-Michel

It turns out that the new version
On Sat, Jul 11, 2015 at 06:20:30PM -0400, Jean-Michel Campin wrote:
> Hi Patrick and others,
> 
> I did not follow closely all the changes from end of last week,
> but it seems that after Friday Jul 03, we have 1 more recomputation warning 
> in several test experiments: 
>  1D_ocean_ice_column 
>  bottom_ctrl_5x5
>  global_ocean_ebm
>  lab_sea
>  OpenAD
>  tutorial_tracer_adjsens
> Is it a benign one that can be safely ignored ?
> 
> Also, it seems that the latest TAF version (2.9.2) that was tested for the 
> first time on Thursday Jul 09 is changing the AD gradient of lab_sea experiment:
> On baudelaire with gfortran:
> version 2.9.2:
> < Thu Jul 9 00:15:11 EDT 2015
> < Y Y Y Y 16> 6<16 FAIL  lab_sea  (e=0, w=8)
> ---
> version 2.9.1:
> > Wed Jul 8 00:15:12 EDT 2015
> > Y Y Y Y 16>16<16 pass  lab_sea  (e=0, w=8)
> Should we update the output_adm.txt ?
> 
> Regarding the OpenAD tests on baudelaire, the 2 recently added experiments
> are failing:
> > Y Y Y N .. .. .. N/O   OpenAD.ggl90
> > Y Y Y Y  1> 0< 0 FAIL  OpenAD.kpp
> 
> And finally, have not seem yet a single halfpipe_streamice test passing
> this month. Does someone care about halfpipe_streamice testing ?
> 
> Cheers,
> Jean-Michel
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel