This is much too long for completely different files.

Apart from the fact that `Global_1993_2021` was obtained with a more
recent version of the code, there is not much point in keeping a
separate experience for the year 1993.

This is useful because, for the year 2021, the last part of Aviso file
names varies.

Define it inside the function since we only need it here and we
already have the argument `first_file`. Simpler and clearer.

Clearer and simpler. Also it avoids appending to a previous
`timings.txt` file.

The bug comes from commit 1cfed454.

`n_max_cont` is just the maximum number of contours in `cont_list`,
and `n_cont` is the actual number.

Create `cropped.nc` and `unpacked.nc` in the slice directory, so that
several processes running `inst_eddies.py` for different slices can
run in parallel.

Instead of symbolic links. We want to avoid creating symbolic links
above SHPC because we want to run several instances of `inst_eddies`
in parallel.

So that several processes running `inst_eddies` for different slices
can run in parallel.

So that several processes running `inst_eddies.py` for different
slices can run in parallel.

So the script can write to a slice other than 0.

On second thought, I find the default value confusing. We are
reverting commit b328ab6f.

The attribute written was always true.

Otherwise we get a harder to understand error from numpy.

This is useful is an SHPC is missing an orientation.

Bug fix: do not provide standard input for `inst_eddies_Aviso.py`, it
does not read it.

Use out_cont%closed instead.

Because stat is a module from Python standard library.

Allowing the module to be imported and used conveniently from Python.

This avoids copying `cont_list(n_cont)` to `out_cont` if it has
insufficient area.