So this removes the sqlite files from version and it will recreate the baseline.sqlite once on each run so the first test will always appear slower than it is because it is seeding the database.
Also because the IntegrationTestCase uses the Orchestra/TestCase and the FeatureTestCase uses the Orchestra/BrowserKit/TestCase that means that each of those recreates the baseline.sqline once for each.
I don't think the time penalty of a second per run is that bad for consistent test runs.
Its hard to explain clearly but I hope you understand.
So this removes the sqlite files from version and it will recreate the
baseline.sqlite
once on eachrun
so the first test will always appear slower than it is because it is seeding the database.Also because the
IntegrationTestCase
uses theOrchestra/TestCase
and theFeatureTestCase
uses theOrchestra/BrowserKit/TestCase
that means that each of those recreates thebaseline.sqline
once for each.I don't think the time penalty of a second per run is that bad for consistent test runs.
Its hard to explain clearly but I hope you understand.