Hi Reece, > (Vista does not set the last error on success for the functions used > in protectdata). In that case, it seems to me we should just remove those tests, as it's unlikely an app can depend on the last error being set on success. --Juan