* Added code to find drivers that are externally loaded
* Support wait_until in driver
* Optionally make xpath absolute paths absolute
* Add predicate to test if drivers shortcircuit waits and tweak spec
* Allows file:/// (envjs) urls as well as http://foo urls (rack)
* Warm the driver in wait_until tests so the init cost isn't included in the test
- Replaced usage of wait_for with locate.
- wait_until back to raising error instead of returning false (This is key)
- wait_for_condition removed
Capybara::Session with culerity driver#wait_until
- should wait for block to return true
- should raise Capybara::TimeoutError if block doesn't return true within timeout
- should accept custom timeout in seconds
- should default to Capybara.default_wait_time before timeout
Capybara::Session with culerity driver#locate
- should find the first element using the given locator
- should raise ElementNotFound with specified fail message if nothing was found
- should accept an XPath instance and respect the order of paths
Capybara::Session with culerity driver#locate within a scope
- should find the first element using the given locator
Capybara::Session with culerity driver#locate
- should wait for asynchronous load
Somehow, using Timeout was causing the Broken Pipe
errors in Culerity. Until someone can explain
how and why, this approach seems to work without
any significant drawbacks.