Capybara tests always returning true

Hi guys,

I am using a test setup with rspec (2.8.0) , capybara (2.0.2),
selenium-webdriver (2.27.2), rails 3.0.20 and ruby 1.8.7

My problem is that every request test I have doing something like:

page.should have_content(“random string”)

is returning always true, even though the page has no random string.

I have tried changing the driver for js that capybara uses from selenium
to
rack_test, but I dont even run the test with the js => true option. But
I
saw on some post that it could help. but it did not.

I am feeling a little lost since everytime I found a post with a similar
problem it was do to some other gem. But I am working in a big project
and
can’t really uninstall and reinstall a gem one by one, to see this.
Anybody
has experienced the same?

all the best,

Andre