I've asked Mercury several times if they have a "Best Practices" document etc. I am gettin to a large bed of test scripts, and am trying to "standardize" them. Right now, I have a main test that opens browser, logs into my site, then runs a series of tests. It then logs out, logs in w/ another userID to run some tests against the functionality of that ID, then logs out.
As I look at Test Director, I'm assuming it would be better to have each script independent from another - ie. have each script open browser, log in, do it's function, then logout. Is this a safe assumption?
I'm just trying to get my test script bed as easy to use/reuse as possible.
I think best practice is not something defined. It is what you follow according to your project & convenience.
The way you described, seems like you are running through TD a serious of scripts. If you are running through TD then you can Create the
1. Name the Scenarios of the Functionality you want to test.
2. Select the respective tests, which needs to be executed to test that functionality.
Basically you are making shore that you cover all the functionality through Scenarios. This would save your time by adding everytime adding the scripts for every Run.