The confusing part was inconsitent reporting. It reported un eaven indices when only one source and a fail, but not when several sources and some success and some fail.