> It's the other way around. The RFC says, > > "Each chained logical bitstream MUST have a unique serial > number within the scope of the physical bitstream." Oh. I hope I haven't made the wrong assumption in too much code :/ Thanks for the correction.