CI: have a step that prevents regressions wrt passing tests

[wprzytula]

Once we have some tests passing, we don't want to lose them after some changes. Hence, it makes sense to introduce a CI step that prevents regressions, i.e. rejects a PR that makes some tests which were previously passing fail.

[adespawn]

Considering, that currently some tests pass because of datastax driver code, I don't thing it's a good idea to add this (for now)

[wprzytula]

Considering, that currently some tests pass because of datastax driver code, I don't thing it's a good idea to add this (for now)

Agreed. Let's keep it in mind, though, to take advantage of regressions detection as soon as possible.

[wprzytula]

Ref: #43