We are going to use entry_points and will need `setuptools` for that.

The former default ``1`` prevented differed from handle_options()
default (``0``). We are back at 100% test coverage now.

We now use an `argparser` options instance for get_passphrase
instead of a bunch of single args. This makes the main function
`get_passphrase` less readable (you do not see immediately what
options are really expected), but avoids misleading defaults. We
now use the defaults set in argparser options, i.e. from
`handle_options()` as single source of defaults.

The `diceware` run in travis was meant as a packaging test. It was
supposed to fail if not all data files were available in
installs. Unfortunately it did not fail in cases where `pip` *did*
fail.

Furthermore we want to detect such packaging problems in regular
tests, before pushing changes.

Move wordlists into the new package. This should fix
the data_files problem on most platforms and in most
environments.

Seems to be really tricky to install data files in a
foreseeable location.

The more finegrained solution of conorsch might be
slightly less stable in the long run, therefore we
selected the 'package', actually.

Looks like readthedocs does not like 'classic' theme.

As we do not install extra packages we do not have to wait for
`apt-get update`.

It is more efficient to run only one test for Python 2.x and one for
Python 3.x. The full set of tests, including all supported versions
listed in `setup.py` should be run locally at least before
releases. Then we also do coverage tests, pep8 and pyflakes.

The 'run diceware' test should remain until we have some other
regression test for bug #1.