[MITgcm-devel] to readme or not to readme

[Martin Losch]

Hi,
I just noticed (or rather was reminded by having a closer look at 
"them") that many of the README files in the verification experiments 
have either little to do with the actual content of the directories or 
the instruction in how to build the model are outdated:
verification/README (incomplete list)
verification/aim.5l_LatLon/README
verification/exp0/README
verification/exp1/README
verification/exp3/README
verification/exp4/README
verification/exp5/README
verification/flt_example/README
verification/front_relax/README
verification/global_ocean.90x40x15/README
verification/global_ocean.cubed32x32x30/README
verification/global_ocean_pressure/README
verification/hs94.1x64x5/README
verification/internal_wave/README
verification/inverted_barometer/README
verification/natl_box/README

Some of the README's belong to experiments, that are no longer tested. 
For some of the README's I am actually responsible myself, but before I 
update those, is there a general idea of what should be contained in 
them (i.e.., a good example of what a README should look like)? I guess 
compile instructions can be found on the manual-web pages and should 
therefore not be repeated in every directory? I personally feel, that 
is it more important to describe, what is actually happening in the 
specific experiment.

Martin