Currently the browser opens and closes between each test method run. Closing the browser requires reopening it for the next method, navigating to the base_url, and logging in. Sometimes this can double the length of a test and for sure adds at least 3-5 seconds per test method in each test instance.
I suggest we launch 1 browser instance per test and make the base_url accessible. This way, a user can define their own setup/teardown behavior without us making assumptions and still guarantee the browser is being used efficiently.
Currently the browser opens and closes between each test method run. Closing the browser requires reopening it for the next method, navigating to the base_url, and logging in. Sometimes this can double the length of a test and for sure adds at least 3-5 seconds per test method in each test instance.
I suggest we launch 1 browser instance per test and make the base_url accessible. This way, a user can define their own setup/teardown behavior without us making assumptions and still guarantee the browser is being used efficiently.