We are thinking about purchasing TD.
In the process of the evaluation we started to think about the "right" way to build our regression in the TD.

Are there any recommendations? Should we have one big repository? How do we manage different sources of the same case for different product versions?