Hi all,

I have searched the forum for advice on how to organize test plans, but have not yet found the information I was after. Maybe it's too much a newbie's questions, but anyway I give it a try.

I want to use TestManager and Robot for testing our application. We will have to test versions under development as well as old versions, when there are patches (every 4 weeks). We do net use RUP.

Currently we have one test plan, that is structured by means of the functionality of the product. We intend to map the tests to be executed within according test suites (which actually can not be the purpose of suites?).

I can not see too much the benefit of the Iterations-folder in the Planning pane of TestManager. As far as I understand it, you can only assign test cases to a specific iteration, by assigning the specific iteration to all those test cases. This means, that firstly it is not visible, which test cases belong to a specific iteration (or at least I haven't found out, how to do this), and secondly this means a fixed assignment and it is not meaningful to use a test case in an Iteration1 and later once again as regression test in an Iteration2 - or I don't understand the feature.

Furthermore, I need advice on how to handle testing artefacts for different versions of the AUT. It seems to be logical to copy a test plan into a test plan for a new version as an initialization step. This would produce new physical copies of - as I believe - all 'TestManager objects' of a test plan, but the Robot-scripts, that are only referenced, would not be copied, but would remain the same. Hence all later changes in the script would be effective for all test plans.

Alternatively we could initialize the testing assets for a new version of the AUT by creating a new project out of the old project. This would also copy the Robot scripts.

I can not yet judge, what the better approach is, nor what alternatives I have. Especially with respect to configuration management the 2nd approach may be nasty, which brings me to my next problem.

According to what I heard from Rational a Rational project can not fully be version controled by ClearCase. As I understood it, it is possible to baseline a project - but only in the sense of being able to reconstruct a specific revision of it. What can not be applied - in my reception - is creating various streams and merging Robot code from one stream to another. I believe, that the configuration management issue is linked to how to perform, when creating testing assets for a new version of the AUT.

Thanks for reading all this. I would very much appreciate, if someonme could give me some advice or point me to addresses with the information I am missing.