Re: [rspec-users] RSpec not failing when requesting missing method/view

2009-06-27 Thread mBread
David Chelimsky-2 wrote: > > That is by design. RSpec is about spec'ing things in isolation, > whereas cucumber is about spec'ing things end to end. RSpec controller > specs are about *controllers*, not views, so the presence and/or > validity of a view should not impact the controller spec. >

[rspec-users] RSpec not failing when requesting missing method/view

2009-06-27 Thread mBread
I've got this test: describe LoginController do describe "GET index" do it "should be successful" do get 'index' response.should be_success end end end which passes, but a cucumber test fails on trying to get the index for LoginController, with the message: No action r

[rspec-users] Asynchronous requests (effectively in multiple browser windows)

2009-06-23 Thread mBread
to logged in users, log out in a different window, then submit the form in the original window. is this possible, and should it be in cucumber that I am testing this, or is it a lower-level test? Thanks, mBread -- View this message in context: http://www.nabble.com/Asynchronous-requests-%28effec

[rspec-users] [Cucumber][Webrat] Using LinkLocator in cucumber step definitions

2009-06-23 Thread mBread
I am wanting to specify "Then I should not see a "New item" link" in a cucumber scenario. Is it possible to use Webrat::Locators.find_link (as used in click_link) in a cucumber step definition? Or is there some reason why I shouldn't assert the absence of a link? I could just use response.should_