Task local variables? Design question?

[jbaker_softcomms]

Maybe I have missed something but I would like to use persistent variables that are local to a task. Is this possible currently?

The purpose of this is say I have a "checklist" task which has bunch of todo items. When these are all set the task should be complete and signal execution to resume.

The problem is there may be many of these checklist tasks within a single process graph on different nodes. I would like to save each item state as part of the task instance but NOT have this overwrite process variables of the same name or variables from other task instances that may also have the same name.

I could also have a Task for each check item but I still am concerned from looking at the docs and code that these values get written back and forth to the token data which it is not really related to..

If this is so and I haven't misunderstood why is this so? It seems to me that tasks carried out by users would have data often associated with that task alone.

[kukeltje]

in the 3.1 tree the taskvariables are local, meaning they are assigned a value from the variables of the same name in the parent token. They will however, when finishing the task, be updated in the original variables and overwrite those values.

If they are so specific to a task, you should consider not storing them in the process engine but outside it and store references to these in the process engine