There is also a prospect of having a "research" workspace, where you have a large code base loaded to search through the code via Eclipse, using Open Type (Ctrl+Shift+T), etc.

If the base table to which the data is loaded from the external table is defined to be a RAW table without indexes, the IDS server USES light appends to load the data.

Usually this isn't an issue, since it only USES base functionality that will probably already be loaded or would have been loaded soon enough anyway.