When executing files it's dangerous to simply assume that the default run configuration is the same as the one you used last time. When running on Beam engines like DataFlow you risk running a big pipeline on DataFlow only to default to that for the next pipeline which you might want to run locally.
To prevent these mistakes we propose to do two things:
- Remember the last run configuration name per file
- Visually inform the user if the proposed run configuration name is different from what was used in the last (different) file you executed.