While working on one of the first major issues, and number of small annoyances have been popped-up in relation to test configuration and behaviour.

The following problems have been observed:

• Running with pytest-flake8 (--flake8 option) will result in tests being shown as skipped. In particular, pytest-flake8 will skip performing linting tests against test files themselves.
• Running with pytest-flake8 (--flake8 option) will result in a lot of noise during development due to styling issues being reported. Although useful in terms of keeping the code clean, it does get annoying after a while with some temporary imports and such.
• pytest-flake8 ignores the .flake8 configuration file, resulting in linting issues being reported for the maximum line length.
• Running with current configuration for coverage ends-up creating the coverage directory in a bunch of directories. This should most likely not be done, especially given that test coverage should be pretty high given the TDD-oriented development.
• In order to see coverage issues, it is necessary to open HTML page. This is an additional step, and could be relatively easily missed.
• Running tests against just the functional tests will produce invalid coverage report, since functional tests are run from a black-box perspective.
• Running tests using tox will fail because coverage will report failure to produce report while running the functional tests.
• Invoking pytest will run functional tests as well, which can be a bit slow, and documentation implies the functional tests are not being run automatically.

The following should be done:

• Remove use of pytest-flake8. This includes configuration, dependencies, and documentation. Keep the .flake8 configuration file, though.
• Update development documentation with instructions on how to run the linter (flake8) directly instead.
• Remove generation of coverage report in HTML format. Instead replace the default invocation to use --cov-report=term-missing:skip-covered. This should provide immediate output and info about what lines are not being covered by tests.
• Remove coverage runs from default target, and have it as an explicit step. Do run the coverage tests via tox, though.
• Update the tox command to run both unit and functional tests using a single command.
• Update the pytest.ini configuration to explicitly list just the unit tests path (tests/), so that invocation of pytest would run just the unit tests (without functional tests).
• Update the development instructions to let the developers know about the behaviour implemented.
• Update tox configuration to ensure the Sphinx documentation does not get built in cloned git directory.
• Update coverage configuration to fail on anything below 100%.