[MITgcm-devel] changes

[Jean-Michel Campin]

Hi,

3 things:

1 ) new tutorial dir.
Following the documentation-session discussion,
I am about to check-in the tutorial dir in MITgcm/verification
as a 2nd copy of the corresponding test exp., and will remove the
old test in the next few days (jut the time to get enough successful 
tests on various platforms).
New name:
  tutorial_barotropic_gyre         - from R1,  exp0
  tutorial_baroclinic_gyre         - from R1,  exp1
  tutorial_global_oce_latlon      <- empty for now
  tutorial_global_oce_biogeo       - dic_example
  tutorial_deep_convection        <- from R1 almost like exp5 <- needs some work
  tutorial_plume_on_slope          - from R1, plume_on_slope
  tutorial_tracer_adjsens          - from R1, carbon(input_ad)
  tutorial_global_oce_optim       <- empty for now (David ?)
I propose to add also:
+ tutorial_global_oce_in_p         - global_ocean_pressure
(which was not mentioned earlier, but described in the manual)

I will add a simple README file (that we can change later on)
that mentioned the old test exp name

  tutorial_held_suarez_cs & tutorial_cfc_offline are already there.
And those that are left on the side for now, are:
  tutorial_offline                 - Helen ? cfc_offline ?
  tutorial_eddying_box_obcs        - Martin!  ?
  tutorial_aqua_coupled            - Daniel  ?
  tutorial_lab_assim               - Andy   rotating_tank

2 ) remove .cvsignore
I propose (and Ed and Patrick agree) to reduce the number
of .cvsignore files:
In each verification/exp_dir:
- keep the one in /build unchanged
- simplify the one in the exp_dir directory to just:
> tr_run.*
- and remove code/.cvsignore , input/.cvsignore , results/.cvsignore
  that are not needed.

3 ) how to download the code using cvs (manual, section 3.2.1,
and from the front page, "Source Code": http://mitgcm.org/source_code.html)
I propose to add the -P option:
cvs co -P MITgcm (instead of just cvs co MITgcm)

I always use it, and I think it's better (specially for a new user)
to get a clean copy of the code without the (potentially confusing)
set of empty dir. Objections ?

Cheers,
Jean-Michel