On 13-06-09 12:47, Eric Adler wrote:
TV that was made in the mid-to-late 1980s (Sony Watchman) and now they are trying to 'solve' the mobile delivery 'problem'.
Qualcomm is working on adding "broadcast" capabilities to LTE (this was from recent at a conference (Telecom Summit in Toronto), so I suspect they would implement as multicast). Right now, there is negative incentive for large ISPs to deploy multicasting on their ISP service because the large ISPs are also legacy TV distributors (aka: cable TV) and that business is highly profitable and they aren't about to help lower cost competitors eat into their cable TV business. However, once internet distribution realy takes off, you'll find the motivation to deploy multicasting will grow because ISPs will want to cut their costs (and that may also be the big incentive to really move to IPv6). But when you think about it, the only time multicast becomes useful is for live broadcasts (sports, sometimes news). For the rest, the world is moving to on-demand viewing where multicast does not provide any advantage. And something to consider: while legacy TV is on /24/365, there is not enough programming to fill all the time, hence the many reruns. Same with infotainment networks like CNN who reruns their stories multiple times per hour throughout the day. In an on demand world, the bandwidth will be relative to the amount of content actually being produced, not the number of hours per day. If you have already seen a CNN reprt on its web site, you're not going to watch it again 5 times during the day. But if you are watching CNN on linear TV, you have to keep watching just in case there is something new that is shown. I predict big changes in viewing habits. And this would have implications on how your guys architect your networks. What used to be TV Networks with stations in every city is likely to become cache servers distributed in every city. (some would call them CDNs :-)