Workspace ids are generated from information that identifies the data that is being sampled. In a file system, it's generated from the file path. This means there is always exactly one workspace for the same file. This means multiple people can't use data prep on the same file without overwriting each other's changes. This also causes issues when pipelines are created from workspaces, as there can be conflicting state within the pipeline and the workspace.
This requires a re-design of workspaces from a UX perspective as well as UI and backend changes