Thanks for your bug report. This was actually an intentional change. Some scripters were decoding in chunks, which resulted in broken UTF-8 text. So newer versions of $decode() do not pay attention to the N. I will update the help file to reflect this.