Scheduler/Process extended return value

[Walter.Koller]

Currently tasks only provide the status: completed / failed that can be used in Schedules to control the process flow.
Branching the execution path can only be achieved by creating artificial termination / fail status. 

It would be nice if the Scheduler can fetch the process return value (output variable) and then apply simple comparison.

[PvD_SE]

Hi Walter,
Two questions:

• What sort of comparison do you have in mind?
• What would this proposal add over the end status and the available log?

Regards,
Paul
Sweden

[Walter.Koller]

comparison like: =, <, >, <>  
or: TRUE, FALSE
Actually it does not matter. The more flexibility the better but everything is better than having no decisions in Scheduler execution.

Currently there is only 'completed' or 'failed' that can be used for evaluation.
The issue is, artificial 'failed' cannot be distinguished from actual failed executions.
And this is not very 'nice' in a couple of ways:

• Monitoring SW will count those 'failed' as issues although they might be intentional
• There is no way to see if there was an actual error during execution (I would not want to check logs every day)
• The process completes successfully, then it should have status 'completed'. An example: If I want to distinguish two pairs of variables of green & red and 1 & 2. It would be awkward to merge them together to : (green && 1) = green ; !(green && 1) = red

In some cases it might be possible to create another processes that does the evaluation and then starts the other processes as sub-processes instead of them being tasks in a schedule. Unfortunately, those sub-processes are not shown in the Control screen and any errors have to be first investigated instead of being visible at a first glance.
Also having this would not work when login / logoff is involved in the Schedule.