John C. Dvorak
The noisiest buzz in the industry lately has been over the emerging use of cable TV systems to provide fast network data transmissions using a device called a cable modem. But the likelihood of this technology succeeding is zilch. It's one of those interesting-sounding ideas that will attract what venture capitalists call dumb money. Unfortunately, it's a big distraction in a market that should be concentrating on ISDN and broadband.
HP, U.S. Robotics, ZDS, and others have been toying with the idea for a few years, and Motorola Multimedia Group's recently announced CyberSurfer 10-Mbps cable modem has completely muddied the waters. There's also LANCity's announced $595 model. Until recently, these things cost a ridiculous $5,000.
Cable modems are, of course, targeted at Net surfers. According to the press announcement, the CyberSurfer will be the fastest, receiving data at 10 Mbps and sending it at 768 Kbps. Exactly how the modem will work on the largely one-way cable systems in the U.S. is a mystery. And since there's no governing standards organization for cable modems, these devices won't be able to talk to modems made by anyone else. But hey, they sure are fast.
Even so, users with access to a T1 phone connection will soon discover that the fastest provider can send data at only around 56 Kbps--slower than a single B-channel over ISDN. This isn't likely to change as providers try to serve as many users as possible, rather than pump 128 Kbps or more to a few people. So the ideal connection for Net surfing is a single B-channel on an ISDN line.
So even if you had a 10-Mbps cable connection, it would be useless except for point-to-point transfer at Motorola's upstream speed of 768 Kbps. And that assumes upstream capability, which the cable companies will have available in only a few test areas. If users don't flock to this technology in those areas, the cable companies will drop the idea like a hot potato.
We have to remember that, collectively, the cable TV folks are as dumb as fireplugs. There is no incentive to be otherwise. They have monopolies and do little more than string wire and milk the cash cow. Why would they want to get mixed up in something that requires real work? If you doubt this, visit your local cable company and ask "When will you have cable modem capability?" Just see what they say. My guess: "Huh? What's a modem?"
Then there is the issue of security. The cable TV system is a broadcast medium, not a secure network. All transmissions over cable are highly susceptible to hacking. Much more so than anything else except cellular phones. HP is one company that harps on the security issues regarding cable modems.
So why spin our wheels over a dead-end technology when ISDN is here now? Is speed really the issue?
There may be another element at play. When you consider digital phone networks and the equipment that is needed to hook up ISDN, you see an interesting phenomenon. The modem companies aren't in the game. Networking companies run the show: Ascend, Cisco, and Combinet. Modem companies like Hayes--and recently Zyxel--miss the mark with ISDN. Others have ignored it completely.
In a digital world, you won't need to MOdulate/ DEModulate (the root meaning of modem). Many users just can't make the transition to a future where the modem is moot. For these sentimentalists, the cable modem is the last gasp. But there are no cable modem standards whatsoever, and very little cable modem promotion within the brain-dead cable TV industry. While this fiasco unfolds, we hear cheering from people who should know better and examine the simple illogic of the whole thing.
Hey, but it sure is fast.
Why do we spin our wheels over a dead-end technology when ISDN is here now? Is speed really the issue?
(Figure not available)
Illustration by James Steinberg
Copyright (c) 1997 Ziff-Davis Inc.