You have been recruited as a TAE by a large travel agency and are now...

You have been recruited as a TAE by a large travel agency and are now responsible for a mature TAS that is dedicated to functional testing of the company’s booking system, which covers flights, hotels, and packaged holidays. This application contains a complex UI and a lot of functionality. It is undergoing constant maintenance and enhancement by an Agile team that adds tests to the regression test suite at the end of every iteration.

The TAS supports structured scripting and this has been implemented to a high standard, with a well-managed script library and minimal code duplication. The TAA will not support any more advanced scripting techniques. The test automation that has been achieved so far was well received at first but is now the target of increasing criticism. It is perceived as being old-fashioned, poorly documented and, in particular:

1. The test script maintenance overhead is becoming excessive because the test analysts frequently want data values to be changed.

2. Test analysts are complaining that a new test which is similar to an existing one, except for data values, requires a new high-level script, and that they must ask a technical test analyst to do this for them.

3. The nightly full regression run is taking a long time and there is a risk that, as more tests are added, it will soon take too long to run in the available window.

4. Some tests break frequently because of slow response from the SUT.

5. Some transactions are difficult to verify because they rely on variables that cannot be seen through the UI of the SUT.

6. The UI of the TAF is difficult to learn.

The SUT architect is planning a major new release and they have offered to include features that will enhance its testability. You already have plans to deal with items 1, 2 and 3 on the list, so now you are considering how you might mitigate items 4, 5 and 6.

Which TWO improvements will BEST address items 4, 5 and 6?

a) Add an event mechanism to the SUT which the TAF can subscribe to

b) Adopt a set of standard setup and teardown methods

c) Add an API to the SUT that will make the internal variables observable

d) Update and improve the documentation of the TAA

e) Implement hard-coded waits in scripts that have been affected by slow SUT response