RFT / UFT more stable??
We're using UFT to make test cases and everything is going well. Also, we are using Pegasus to control multiple tests together. The problem is that some times one test stops with no obvious reasons and it affects other tests.
When we try to look deeper why the test stopped, we find something like UFT wasn't able to detect a certain window (although the window is already open and UFT always detects it except this time)
Do you think if we use RFT, will it be more stable? Stable in the sense that if it detects something, it will be able to detect it again as long as same conditions.
The problem you are talking about should be handled by different synchronization mechanisms of your own. To control your execution and scripts, you have to have some structured framework kind of setup and good sync methods utilizing the available UFT/RFT features like Wait, Exist, WaitProperty, etc.
Also sometimes it is identifying and sometimes not may be not enough unique properties to find. Help UFT/RFT by providing right set of unique properties to handle the object identification smooth.
Thank/Like to help others if my input helped you !!!
I am speaking for myself, not for my employer nor any one.
In my experience, RFT has a higher resource overhead. Most likely given everything is the same, UFT probably would be more stable.
But as the previous person mentioned. Tests are rarely stable without a good amount of effort invested. You pretty much have to treat everything as aync code and make no assumptions about the element being there or not. Wait wrappers, fall backs, etc.. are pretty common in test code. You tend to have to write a pretty large library of these.
Does it look as if your issues are sync related in UFT? I have been having trouble with UFT 12.02 where it just locks up. It seems to have something to do with the debug tracking.