The %{error.taskindex}, The position of the task in the process is only available if there is an error. It would be nice to be able to access this value and other error only variables even when there isn’t an error.
%{error.tasktype} - The type of task that triggered the error
%{error.taskname} - The name of the task that triggered the error
%{error.taskindex} - The position of the task in the process
To reduce logging it would also be nice to set minimal logging on a process, but have a Level 5 for scripts that would overide the setting and log even when the process is set to not log on success.
Can you let us know please what you mean by “a Level 5 for scripts”? The reason why I am asking this is because as in regards to the function Watch.Log() I count four levels only.
Can you also let us know please if checking the option Minimal logs on success in the Properties window of the specific Workflow process is not an option?
You are correct that there is only Four (4) levels currently. The feature request would be to add a Fifth level or some other mechanism where we could script a message to log even if the process has minimal logging set.
Watch.Log("this would log even if minimal logging was enabled", 5);
There are times when minimal logging would be great, but we need some very limited logging of specific values for verification or troubleshooting, but don’t need all of the extra logging. For example just some of the values of variables like the customer number, check number, division, etc. Not all of the process stuff, just specific values.
Both suggestions make sense. The one about logging is a little more involved, however, as the logging process is centralized and has a progressive structure (higher logging levels always log messages with a lower logging level). But that’s for our architects to resolve!
Having access to the taskindex outside of error would be handy especially due to there being no easily assessable run log.
We’ve had to build our own logging subprocess & web front end for a run log but have to manually add the taskindex into each call. These then become out of sync when tasks & branches are added before our log call.