Andrew de Quincey wrote: > On Thursday 16 June 2005 12:45, Johannes Stezenbach wrote: >>On Thu, Jun 16, 2005 at 02:10:10PM +0400, Manu Abraham wrote: >>>Is service_id supposed to be unique in a bouquet too ? or is it only in >>>a transponder ? >>ETSI TR 101 211 says (4.1.1 Network Information Table (NIT) information): >> >>"The SI uses two labels related to the concept of a delivery system, >>namely the network_id and the original_network_id. >>The latter is intended to support the unique identification of a >>service, contained in a TS, even if that TS has been >>transferred to another delivery system than the delivery system where it >>originated. A TS can be uniquely referenced >>through the path original_network_id/transport_stream_id. >>A service can be uniquely referenced through the path >>original_network_id/transport_stream_id/service_id. The >>network_id, thus, is not part of this path. >> >>In addition each service_id shall be unique within each >>original_network_id. When a service (contained inside a TS) is >>transferred to another delivery system, only the network_id changes, >>whereas the original_network_id remains >>unaffected." >> >>However, in practise you cannot rely on this. > > Is it worth supporting bouquets in the file formats do you think? None of the I think if we were to support Bouquets, the config would be become larger, but we can have them cached in memory only rather than write out to file .. ? What do you think ? Manu
