In message <20030708092651.GM4808(a)stuckist.macewan.gen.nz>, Hamish MacEwan writes:

On Sat, Jun 28, 2003 at 13:05 +1200, Ewen McNeill wrote:

...

In message <5.2.0.9.0.20030627160856.02023200(a)pele.citylink.co.nz>, Richard Naylor writes:

...

ah - its that word "channel" that I don't see in the future. Such a dated concept.

Ah. Channels are double-plus-ungood. Call it a stream if it makes you feel happier. Or VDC or whatever.

I think Rich's point is broader than "channels," it goes to the notion of "broadcast," which, while like most things is not going to go away anytime soon, but the option of obtaining material from P2P or other sources (Tivo as so aptly observed by Joe) and assembling your own experience is becoming more feasible,

Except that a Tivo listens to (a bunch of) broadcast channels (there, two horrible words in a row!) and caches the "interesting" stuff that it sees for later viewing. The problem, IMHO, is not the broadcasting, or the channels, themselves. It's how people make use of them. Broadcasting is a fairly efficient technique to get information to a whole bunch of people in one go (for them to use now or later). Channels are an efficient way to know which broadcast to listen to for what things. I simply don't buy the argument that says that because the "must watch it when it screens, must watch what they want to show" model of broadcast channels is bad (which I definitely agree with -- I haven't even had a TV for 5+ years), that means that all the components of it (broadcasting, channels, etc) are totally without merit. To get, vaguely, back on topic -- multicasting is just "broadcasting with focus" -- if you like it's a more efficient form of broadcasting for a medium which can handle sending only to interested parties (eg, a packet switched network like the Internet, as opposed to a frequency division "network" like radio waves).

[Spoonfeeding, etc] This is what makes me leery of attempts to make the Internet a better broadcast medium than it already isn't. It may make non-broadcast use more difficult. Unintended consequences, increased core complexity, et al.

I'm not especially convinced by this argument either. There's a whole lot of money behind turning the Internet into a "content provision" network -- and I suspect they'll succeed in doing what they want to do irrespective of whether we try to make it more difficult. It does, however, seem to me rather counter productive to insist that they must use unicast, must use the bandwidth inefficiently, in order to some how deter this. At best all that's likely to achieve is that there's even less bandwidth available for other things. (Witness allocations of the frequency spectrum to radio, TV, etc, and what's left for ham radio operators.) And to make it even harder for "small operators" to join in -- so bang goes your online equivilent of public access TV (it's just too expensive to get a pipe big enough to unicast it to a reasoanble audience). [Choose to watch something, join VDC starting it soon]

Yes, like video rental already provides, without the drive there and back, and PPV on Sky Digital.

Yes, like a video rental. Without having to leave the house in the cold of winter. And without having to deal with the fact that some other bastard has taken out the thing you wanted to watch. And, if done properly, without having to watch it all at once so you can return it in time. (I've never seen how PPV on Sky Digital works, so I couldn't say how it compares.)

PS. Road congestion, like unicast broadcast congestion, seems to arise from the continued belief that we all have to use the road at the same times (morning and night, Monday-Friday)

Indeed. And that survives partly because many people "need" to be at work at the same time, which is largely true because of network effects (eg, you'd like the cafe to be open when you want to get morning tea, and people need to be in the same place at the same time for meetings, etc). And in the same way when something "interesting" happens lots of people are going to "need" to see it at once, if only to keep up with the Jones and be able to talk about it at morning tea the next day. To use your earlier analogy providing more buses/trains/etc can help to reduce this congestion. And in the same way multicasting can help reduce the congestion. But apparently buses/trains/etc are good, and multicasting is bad, for reasons I don't follow. Ewen