So! I need some input on this. I am testing an app with Capybara and Poltergeist. The app has a component mounted as a rack app in it's routes. That component interacts with the main app via a http api the main app provides. That means that a request to the front landing page can result in an "internal" GET request to the main app that is then used to make up the data to display on the landing page. The
problem is that the second "internal" request times out since the web server the tests run against is already busy with the current request. In production this is not a problem since I have multiple processes taking care of the requests. Any idea if it is possible to make poltergeist work multiple requests at the same time while testing?
rushed: not convinced... I run integration tests to test the full stack setup.
So, sketching out an idea here. A pretty complex sql with UNION from two tables ending up as a result set with a "type" col that corresponds the the object type that each row should be loaded in. Feeding that into AR to take advantage of my existing pagination and good stuff there. Am I getting myself into murky waters?