In our Python OSS sync meeting this week, a new "environment template" idea was suggested. A user could create a environment template (say, a requirements.txt plus Python version) and then use that template to create a new workspace. Someone might have a template for Shiny apps, for ML projects, etc.
If the IDE can be a little pushy about this, it can encourage good practices for Python users. Not adopting these kinds of practices are a major problem for Connect users right now, and they get into big trouble when they try to deploy an app or doc or model.
I think this would mostly apply to Python users (much more than R users who have a happier time using one global set of packages), but there could be an analogous workflow for renv users.
In our Python OSS sync meeting this week, a new "environment template" idea was suggested. A user could create a environment template (say, a
requirements.txtplus Python version) and then use that template to create a new workspace. Someone might have a template for Shiny apps, for ML projects, etc.If the IDE can be a little pushy about this, it can encourage good practices for Python users. Not adopting these kinds of practices are a major problem for Connect users right now, and they get into big trouble when they try to deploy an app or doc or model.
I think this would mostly apply to Python users (much more than R users who have a happier time using one global set of packages), but there could be an analogous workflow for renv users.