We're seeing a peculiar problem with RFT/Harvest that I was wondering if anyone who has ever encountered this might be able to offer some insight.
Here's the scenario:
Person A creates a new RFT script and commits it to Harvest.
Person B updates and pulls down the new file.
Person B's RFT sees that file as straight Java code, while Person A's RFT sees it as an RFT script.
That works in reverse, too. Person B creates a file, but Person A effectively can't run it. If person A modifies person B's file and commits it, person B's RFT will then not see the same file as a script anymore.
The icons for the files in the Script Explorer differ, good being a script icon and bad being a Java icon.
It seems like it might not be committing dependencies to the repository… so if RFT can't resolve the Helper class that goes along with a script, it might not be realizing it's an RFT script. I notice it's also not pulling any test objects when checking out a file.
It seems like an RFT environmental issue with it not recognizing the difference between a file that's plain Java code and RFT-generated Java code. When opening one of the files with a "J" icon, RFT displays a message in the Script Explorer stating the "Funtional Test Script Editor is not active." When a file with the "script" icon is open, the helper class, datapool, verification points, and test objects folders are displayed.
I've looked around on this and other message forums and see a few folks that have reported a similar problem, but no one ever responds with a solution.
If anyone has any insight to offer, please respond! Thanks in advance!