At 9:48 AM -0600 2/23/05, John McClenny wrote: First John, thanks for this detailed response. There are many issues here in which we are incomplete agreement. There are some key issues however where further discussion is appropriate. >The Cable IP networks are designed for high speed access with bursty >traffic. It will not stand the load of continous delivery of TV video >to lots of TV sets. The problem isn't so much at the edge, but closer >to the core of the network. They are not currently sized for this >large, continous load, Agreed. But this thread was not specifically about the migration of cable to an all IP infrastructure. I agree that the availability of higher download speeds is not in and of itself a guarantee that they can be used to deliver TV services. But the case for any new IPTV deployments by the telcos is much different than for cable, since they must assume up front the kind of usage statistics that you described in your post. And the cable systems are designed to migrate to the kind of loading factors you talk about - to do this they will need to reduce the number of homes on each local loop, light up more trunks to accommodate the increased traffic, and beef up the pipes that connect outside their walled gardens. The shift that I am talking about is not likely to take place over night. The bulk of the traffic on both Cable and Telco IPTV networks will continue to be in the form of IP multicasts of popular content. The shift will occur first in paid downloads of content, as we are seeing today in the music business. In time it may migrate to Internet TV channels that parallel the Internet radio channels that are proliferating today. Even more important will be the ways in which the major content conglomerates refocus their business to make their content libraries available to the public through both the public Internet, and through walled gardens. This is where the decoupling of content and carriage is gong to play a major role. It is helpful to look at what is happening around the public Internet, with respect to the infrastructure needed to deliver digital media content, and WHO is paying the costs for the required infrastructure. At the most fundamental level, there is a huge difference between the Internet model and the walled garden multichannel TV model. With the Internet model, content and carriage are almost totally decoupled. AOL still has some unique content franchises, but most of the content they offer today is actually coming from the public Internet. With the Internet model, a content owner PAYS for distribution. This may be nothing more than a small monthly fee to an ISP to host a site, or it can be a huge, costly effort. Akamai does not pay content owners for the privilege of delivering their bits - they get paid to do this. It is then up to the content owner to figure out an appropriate business model to make a profit. This may be a pay-per-view fee, a subscription fee, or the content may be supported by advertising revenues. And there is another category that properly considered a marketing and/or support expense. For example, Apple has a popular movie trailer site; they use Akamai to deliver the bits. There is no advertising per se, but the movie trailer site promotes Apple's QuickTime, and there are links on this page to other apple products and services. What is IMPORTANT here is that there are MANY business models that can be supported over the same distribution infrastructure. With the walled garden model the "marketplace" is highly distorted. In many cases the content conglomerates are paid for their content, rather than paying for carriage costs. As these fees are passed through to the subscriber (often with a markup for the Cable MSO) it is easy to say that these are subscriber fees. But subscribers are not given a choice as to whether they want this content, and many would drop some of these channels if they knew how much they were paying for something they rarely if ever watch. Most browser channels would drop their subscriber fees if ala carte became a viable option, because they would lose a big chunk of their audience if people could choose whether to pay for that channel. The parallel with the Internet model is that the content owner would pay the cable company for carriage, not the other way around. >The pricing we pay for broadband is based on relatively small amounts >of data used per month. The BitTorrent users, although few in number >and transfering relatively small amounts of data, are major loads on >ISP networks. To build a network that can sustain even an MSO >equivalent video load with mulitcast and unicast VOD is neither cheap >nor easy. Look at Broadband reports and all of the high data volume >users getting kicked off of cable systems when they hit some usage >caps. Now make everyone a high data volume user and the system >collapses. But this is just a reflection of the approach that cable systems have taken with broadband services. And a reflection of the fact that they do not need to provision their networks to deliver the IP MUlticast bits, since they are broadcasting this stuff to everyone. As the usage statistics evolve for cable they must reduce the number of homes on a local loop and beef up the plumbing all the way back to the core. Any new IPTV deployment assumes that everyone will be a high bandwidth user. The network is provisioned from the start to support the kind of bit volume that would bring a HFC cable network to its knees. So the problem is not near the edge, or in the trunks from the CO, but rather inside the CO and in the links to the outside world. Note I did not say public Internet. The links I am talking about are a combination of public and private networks. For stuff that is available via the public Internet this will require some big pipes, as it does today with DSL. These pipes can be scaled as needed based on usage demands, and the cost will be reflected in the rate structure, or possibly in usage fees based on the actual volume of bits consumed. But there will ALSO be private networks bringing content into the CO, just as there are private networks that bring content to cable head ends. The good news is that the content conglomerates typically pay for the satellite bandwidth used to deliver the networks feeds to the cable head ends. I would expect this to be true for the private network bandwidth that content owners would use to deliver content to IPTV systems as well. For example, Disney would most likely create a private network infrastructure similar to Akamai, or they could choose to pay Akamai or some other competitor to operate the network for them. And they would pay for a big pipe into the IPTV head end so that the IPTV system subscribers could access the servers with the Disney content library. This private network would be used to deliver the stable of Disney IP multicast channels (ABC, the ESPNs, ABC Family Channel, The Disney Channel, et al) . And it would carry the bits for all of the demand based services as well. It is highly likely that Disney would seek to co-locate servers at the IPTV Head ends, as is starting to happen with some cable networks. These servers would mirror popular content to take the load off of the wide are links, and they would handle the localization and personalization of the content. Bottom line, the business model would evolve and it would be much more obvious to the consumer what they are paying for and to whom. >Right now the money paid for content subsidizes the carriage. Both >the MSOs and telcos leveraged existing infrastructure to add IP >connectivity. Moving their networks to the next size level involves a >lot of forklift upgrades if you want to build an infrastructure that >can handle the general any content to any person case. What are we >willing to pay for this? The obvious answer is YES. How many of us on this list are paying for both a broadband service and a multi-channel TV service. And how much do we spend in addition to this, to rent or buy content? I can go to a bunch of local stores and spend $10-15 for a music CD. But now I can also go to Napster, Serius or XM for an "all I can eat" music subscription service, or I can go to the iTunes Music Store and its competitors and buy just the tracks I want, or the entire CD for $8-9. The true cost of the album from the iTunes store is greater than the money paid to Apple, because I need to include a portion of my monthly fee for Internet service, but the savings I get versus going to the store help to offset this. Apparently the Telcos believe that they can deploy an IPTV infrastructure and still be competitive with cable. Moore's law is working in their favor, and there is PLENTY of wide area bandwidth available to deal with the issues of moving bits at the core of the networks. The cable guys just spent about $70 billion to upgrade their networks, and they will continue to spend big bucks to upgrade them as they migrate to an IPTV infrastructure. Their reward is that the revenues per subscriber are increasing as they introduce new service. Will all of this be disruptive? Of course. Will it cost consumers more? Perhaps in the short term, but it will cost less in the long term. Why? Because of competition and the fact that it will be much more difficult to shift revenues based on techno-political regulatory advantage. The marketplace works! >If we separate the carriage from content, we have capped the potential >revenue for the carriage provider while forcing them to bear all of >the risk of building and supporting the network. The business id iffy >enough under the current setup, remove any hope of incremental revenue >and the networks will not get built. Unless, of course, we let the >governments build and subsidize the networks, but that is another The Telcos have been in the carriage business for almost a century. The revenues they produce still dwarf the revenues produce by the entertainment industry here at home. We still spend more on beer as consumers than we do on electronic entertainment; and the beer industry is one of the largest advertisers on TV. What's more relevant is that both cable and the telcos are making money on broadband. There will be MANY ways to generate incremental revenue. There is PLENTY of money in the system. The real issues are the equitable distribution of this money. Even more important is the reality that many independent producers will be able to bypass the gatekeepers. It will be possible to create a series that does not have to be sold to one of the media conglomerates; you can sell it direct to the consumer at a much more attractive price. Competition will help to bring the cost of content under control as well. >I think the reality is that you cannot build a business case for a >truely open network at this time. Too much risk and not enough >revenue. I am not trying to build this business case. I AM predicting that this is the way things will evolve over time. IP multicast tiers will be a necessity for competition with cable in the near term, and they will continue to be necessary for some forms of content in the future. People will still want to watch live sports, news and weather. But the consumption of pre-produced entertainment programming is going to shift from todays browser model into a model that is driven by the consumer, who will become the program director. Adjaceny to a program on NBC will become meaningless. Getting your content into a viewers PVR is going to be the new game. You will surf the stuff you want, not the stuff they want to push at you. And the compensation models are going to change drastically too. Advertisers are going to move their money where they can get verifiable results. Paying for thousands or millions of eyeballs that are NOT paying attention is only going to work in limited applications. The advantage of IPTV routing is that you know EXACTLY what is being delivered to each home. The real value add for IPTV systems will be to leverage this information in ways that are acceptable to both the subscribers and the people who want to reach them. >Ala carte pricing brings more transparency to content pricing. It >would raise rates for, say ESPN, because the base of subscribers would >be smaller. Under the current system, I am subsidizing ESPN viewers >if it is part of the base package and I don't use it. The longer term >advantage of ala carte is reducing the pricing power that the content >providers have right now by allowing consumers to make more >incremental choices. All true. It is the pricing power that they enjoy through regulatory advantage that we are all paying for today. We may want ABC or ESPN, but we must pay for ABC Family and Disney too. Hopefully we are close to the peak in terms of content pricing. Clearly the major sports franchises have peaked. I suspect that the days of an actor making $1 million per episode for a sitcom or drama are numbered too. > >You still have the wall between PC and TV content. It isn't going >away anytime soon. Not sure I understand this. I believe this wall is primarily a difference in applications. Up close and interactive versus lean back. But the lines will blur here too. The same technology that makes the interactive PC experience viable today can be used to localize and personalize the TV experience,. Even more important, TV display technology is finally getting good enough to support the new interactive applications that will be optimized for the big screen in the family room, rather than the PC in the den. Regards Craig ---------------------------------------------------------------------- You can UNSUBSCRIBE from the OpenDTV list in two ways: - Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word unsubscribe in the subject line.