Received: from eff.org by kragar.eff.org with SMTP id AA06116 (5.65c/IDA-1.4.4 for ); Sat, 11 Apr 1992 13:30:40 -0400 Received: by eff.org id AA01535 (5.65c/IDA-1.4.4 for pub-infra-exploder); Sat, 11 Apr 1992 13:03:26 -0400 Date: Sat, 11 Apr 1992 13:03:11 -0400 Message-Id: <199204111703.AA01530@eff.org> From: Mitchell Kapor Subject: Cable Television and the National Public Network To: pub-infra@eff.org (pub-infra mailing list) Following are my postings excerpted from a thread in the Well's EFF conference. Due to Well policy, I cannot supply other posters' contributions. My remakrs stand alone fairly well. I begin by replying to an accusation that EFF has swallowed the telco line on ISDN. EFF has not taken the telco "LINE" about anything, much less that they are the idea provider of telecommunications. What we have said is that ISDN is tactically very interesting as a first step toward a national public network which is digital, open, widely available and affordable. That said (and you can refer to the numerous statements we have made on this subject), it must be added that cable offers some interesting possibilities. It has a high bandwidth (hundreds of megabits per second) and already reaches 60% of homes, passing by over 90%. The biggest issue we see is that while the telephone system operates on a common carrier basis which requires the phone companies to accomodate all comers who wish to sujpply information on the network, the cable system operates under no such obligation. In fact, there is an enormous vertical integration in cable already with the major cable systems such as TCI and Time-Warner owning major interests in cable channels like CNN, HBO, and others. NBC was unable to mount a CNN competitor because TCI and others refused to carry it. In our vision of the national public net, it is crucial that everyone be allowed to participate, not only as an informationm consumers, but as a provider. Common carriage is the way to enable this. Interestingly, there are some preliminary moves suggesting ways in which this might be accomplished. One proposal, filed in the FCC hearing o video dialtone, suggests a "condominium" approach in which cable would install a fiber-coax hybrid system nationally and sell digital carrying capacity t other carriers (LEC's, long-distance carriers, etc.) who would then operate that portion on a common carriage basis. We are interested in exploring these options and discussions are underway. --- 64kb is not the upper bound for transmission over the copper local loop. ADSL and HDSL both offer high bit-rates to the home. ADSL provides full T-1 from the CO and some amount (9.6-64kb) back. HDSL currently operates at 768Kb, but it is fully symmetrical. You would need two pairs to get the full 1.544Mb. Both ADSL and HDSL are transmission protocols. In all probability ther higher level layer of the stack will be adopted from ISDN, according to the folks at Bellcore we spoke to. ISDN should not be thought of as simply providing a 64kb "B" channel, but as a protocol suite which can be extended to operate at higher speeds. In fact, Primnary rate ISDN's bearer channel's operate at 1.544 Mb. So ADSL or HDSL could be the means by which primary rate ISDN is made to run over a single copper pair. By focusing on ISDN, there is in fact a migration path to higher speeds, not a dead end. Basic rate ISDN is being deployed now. ADSL and HDSL are still under development and going into field trials. It will be years before you could get it at home, and that assumes that the telcos will be of a mind to tariff it affordably. We like HDSL because, as a symmetrical system, is will allow users to originate high quality video as well as to receive it. Meanwhile, it is likely that higher speeds and longer distances can be achieved over copper. At a Broadband conference last week, an infrastructure planning manager at Ameritech told me he thinks it's possible to deliver 3-6 megabits/second over the local loop using ATM protocols. Speculative, but enticing. The RBOCs have not given much thought to high- speed transmission over coper until very recently. We think they should pursue these prospects diligently. Various hybrid systems, comvining copper and fiber or copper and coax also seem worth investigating. We are making a visit to the FCC this week and to Cable Labs at the end of the month. This subject will be on the agenda both places. Meanwhile, EFF Cambridge is ordering ISDN lines for the office and at home of staff members. It's available in Mass. and priced at 1.6 cents per minute. We'll let you know how the experiments go. Right now it's only available within individual central offices, so its utility is somewhat limited. But it should enable users at home with Macintoshes to operate like they're on a Localtalk network to the office. --- The telcos have seen ISDN primarily as a voice service, whereas the immediate demand will be as a data service. ISDN adapters are available today for PC's for the same cost as a high-speed modem - $300-$500. Prices will fall further as volume goes up. Sun is widely rumored to be building in ISDN into every workstation. It will just BE THERE. Telecom market research firms may have some of their heads wdged in the same places as some of the telcos. Obvious ISDN applications exist now for LAN-extenders, work at home, Internet at home, etc. Enough to drive the first 100,000 users in the U.S., to show there is real demand for the service. Video telephones will be a very big market for ISDN as consumer units offering good quality over 64kb (bot 56) come to market over the next two years and crash through the $1000 then $500 price points. Cable has interesting possibilities, but they are not here and now. ISDN is being deployed here and now. Coax cable cannot handle two way high quality video now. Architectures to permit this are just being explored now. It will take several years if not a decade to develop the standards, protocols, implementations, and peripheral equipment required. We encourage this but think that's too long to waIt in the absence of an alternative. EFF is interested in a platform which is digital, has wide-spread availability, and which is affordable. Cable systems could play a key role here. We're interested in exploring this with them. Cable reaches over 60% of households and passes by over 90%. It meets the wide-spread availability criterion. Coax is very high-capacity, 1 gigabit over short distances. Hybrid fiber-coax systems, in which trunks are fiber to the pedestal, and existing coax to the home are being investigated heavily by cable industry. This is good. The cable industry will use digital cable to deliver more pay-per-view and video on demand of movies and other entertainment. This will pay for the investment required to upgrade (presumably). Cable itself is not under common carrier regime. This is a problem. Cable should consider creation of digital common carriage pipe within a pipe. Dick Leghorn's condominium scheme (proposed in his filing in the video dial-tone case) represents one approach in this direction. We think it should be explored further. In that approach other carriers like LEC's, IXC's own and operate common carriage service which runs through cable system. There has to be sufficient overall capacity, and new cable systems have to be properly architected from the outset to support this. Of particular concern is making sure it's fully interactive. Existing cable has trouble with interactivity, as architected with tree and branch structure as a one way system. It is not necessary that system be fully symmetrical, but it is necessary that the system allow for origination of high-quality video at any point, not just at the head-end. The cost to originate high-quality video need not be as low as cost to receive, but still needs to be affordable. This has to be defined. Finally, cable should be open to explore more creative relationships with telcos in the area of public infrastructure. For instance, in the use of ISDN in the near-term coupled with one way digital cable. One way digital cable can be done now without much if any enhancement of existing cable (I am told). If coupled with ISDN (meaning the subscriber has to have an ISDN line too) could be powerful next step. What's key here is to incorporate in the set-top converters the necessary electronics for both cable and ISDN in this case.