Message-ID: <01f001bf3b95$fe1f6720$0a8db40a@netchannel> From: "Jack Lang" <email@example.com> To: "Jeff Sussna" <firstname.lastname@example.org>, <email@example.com> Cc: <firstname.lastname@example.org> Date: Wed, 1 Dec 1999 00:49:51 -0000 Subject: Re: TV Broadcast URI Schemes Requirements > > It seems like the one thing we all agree on is that there are two levels of > TV URI: physical (channel/time) and logical (program), and that the two must > be mapped to one another. I don't really see any way of making physical > URI's "universal" without a universal physical transport protocol. IP > defines such a thing in the data world, but the TV world does not (unless > you want to refer to things by their physical location in the > electromagnetic spectrum). Err...I'm not sure I agree. Firstly: Physical is more complex. By channel you may mean service name. Service name space has been discussed before, for example in Dan Zigmond's proposal. Services map in complex ways to physical transmissions, and vary with locality and time. The mapping may not be unique or complete. There is not only the actual means of transmission, of which analogue channel is one example, but may be more complex in a digital world (e.g a bouquet in a MUX). For the same service these may vary according to locality, and possibly time of day, and the service and content may be available from more than one physical transmission mechanism for some locations and times. They may be available on demand as VOD, or via a web-broadcast. Secondly: Time is also complex - do you mean absolute time, local time, Normal Program Time for this station and transmission method etc? What happens when the schedule slips? What happens for programs time-slipped using a DVR? The same content may be received at different times from the same station by different transmission routes, for example in the UK BBC1 is currently transmitted in 27 different variants. Thirdly: I'm not sure why programs must be mapped to channels or physical transmission methods as you state, or even that there is a mapping. They are different entities, and different and orthogonal namespaces. The same program may be transmitted by many different physical transmission methods, and different programs may be transmitted by the same physical method (e.g. analogue Channel 2) at the same time in two different locations. A program may happen to be transmitted in a particular locality via a particular physical method at a particular time, or scheduled at a different times, but might also be stored locally on a disc. An analogy might be a web-page and the particular physical transmission route of the packets. Certainly a statement like "Program X was transmitted on Service Y at local time T and received here via transmission route R" has meaning. An EPG might state locally that Program X is expected to be transmitted on Service Y, which this user prefers to receive via route R, at time T. However X, Y and R and T are all separate entities, and the mapping is not a fixed mapping. Indeed the EPG example may dynamically change as the schedule slips. Jack Lang .