One of Flow’s biggest and most unique capabilities is its ability to let you set up processes where data travels from the output of one Flow element to the input of the next.
To use upstream data, the user clicks on a combobox in a Flow element property editor and selects from the available references.
For some kinds of data, these references are created automatically. For other kinds, you have to do it manually. There’s a third bucket of kinds of data that give you the choice to do it manually or have it created automatically. More detail is available here.
As of Spring ’20, automatic references are generated for all kinds of Flow data except for:
- Subflow outputs (coming in Summer ’20)
- Loop index (the reference to the current member of a collection in a loop, coming in Summer ’20)
- Pause element
Automatic Field Discovery
Automatic Field Discovery causes a Get Records to automatically pull in the necessary field values. You no longer have to manually choose them.
Considerations When Choosing Automatic Vs. Manual
The future is Automatic. Use it when you can. It’s faster and easier.
When setting up Conditional Field Visibility relationships between fields, if the condition depends on the value of a screen component, make sure to use Automatic output handling. That’s because the manual variables don’t get updated until you transition to the next screen.
Currently, if you use Automatic Output Handling with a screen component attribute that you’re passing an input too, Flow will remember the most recent value for that attribute. If you Previous to a point prior to the screen and then return to the screen it will not update to a new value. This is discussed more here. We’re changing the behavior in 226 so that Next transitions will always recalculate the input attribute.