Workspaces inside a grouped workspace can be referenced by their name and as a consequence can be over written inadvertently. For example, workspace ‘sample’ in workspace group A and workspace group B refer to the same workspace. Wouldn’t it make more sense for the grouping of workspaces to behave more like folders than a collection of references where the full path is needed to access the workspace.
This would unfortunatel require quite a lot of rework because several aspects within Mantid fundamentally expect workspace names to be unique. Primarily the Analysis data service, the store that holds all of the workspaces in memory holds each workspace by name, even if it is stored within a workspace group, and any algorithm can be run on any workspace even if it is part of a group, and at the moment it is assumed that the name alone is sufficient for a user to determine what workspace to operate on.
This is not to say that in the future this could not be changed, just that it is not a simple or cheap change to make.
Understood, thanks for the response.