03-19-2007 12:37 PM
03-20-2007 08:28 AM
Hi Hurst,
The main problem here is going to be dealing with all the different step types. Because Numeric Limit Test will pass different information to the parameters than a Pass/Fail Step. You won't see any of the information there during development but if you put a breakpoint in your SequenceFilePostStepFailure callback then you will be able to look at the parameters in the Context tab. You will see that if you fail a Numeric Limit Test you can get access to the limits by using Parameters.Step.Limits.Low/Parameters.Step.Limits.High. However, this wouldn't exist if you had a different step type. By looking at the Context tab when you put in a breakpoint you will be able to look for the other variables as well.
Hope this helps.
03-20-2007 10:24 AM
03-22-2007 01:18 PM
Ok, that was good info. I've got the basic part of the fault isolation function implemented and working great. However, the callback causes a new problem, which I hope some knows how to solve.
I have a custom Operator Interface (OI) based upon the full featured OI that came with TestStand 2.0 (it runs fine under TS 3.0). In this OI I am using the Trace Callback to display a list of executed steps in the execution window. The problem is that when a step failure occurs, both the PostStepFailure and Trace callbacks are called (I haven't determined the order). But when I extract the step information in the Trace callback I find that it is giving me the information from the step being executed by the PostStepFailure Engine Callback. The test report, however, has the correct information for the steps in question so I know that TestStand knows how to do it correctly.
Hurst
03-23-2007 10:52 AM
03-26-2007 10:32 AM