Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
I, Cringely - The Survival of the Nerdiest with Robert X. Cringely
Search I,Cringely:

The Pulpit
The Pulpit

<< [ If You Can't Beat Him, Join Him ]   |  Noorda's Revenge  |   [ It's a Sony ] >>

Weekly Column

Noorda's Revenge: IP Telephony is About to Shakeup What it Means to be a Network Software Company

Status: [CLOSED]
By Robert X. Cringely
bob@cringely.com

Back in 1983, AT&T went to Judge Green and suggested that the nation'sphone company be dismembered. It was all AT&T's idea. And the motivation behind that idea came from AT&T's weariness with a lingering antitrust suit and a desire to make even more money. AT&T's dismemberment plan said that its many local phone companies would be freed — not as a single local phone company, but as a number of Regional Bell Operating Companies. One local phone company might be too powerful a competitor against AT&T. In return for giving up its local presence, AT&T would be freed from both federal litigation and some federal regulation. Specifically, they'd be allowed to keep their most profitable division — long distance — and that division would be far less regulated than in the past. And they'd finally be allowed to enter a whole new industry — computers — where AT&T expected to be incredibly successful. After all, Ma Bell was the mother of Unix.In the short run, AT&T got exactly what they wanted. Long distance, even in a new, more competitive environment, was a big success. The AT&T computer business wasn't a success at all. Their line of minicomputers bombed. Their PCs, primarily from Olivetti, were dogs. And their purchase of NCR to tie it all together ended in a spin-off and a big loss. AT&T could reenter the computer business at anytime, but don't bring it up with them - it's a sore subject.

What AT&T didn't envision back in 1983 was the communication and entertainment environment of today. In 1983, the Internet existed, but was used by only a few thousand people and e-commerce was actually illegal. There was always Compuserve, Genie, The Source, and the nation was riddled with local bulletin board systems, but the online world was occupied by only a few tens of thousands of people, and those were communicating at 1200 bits-per-second, tops.

In 1983, the cable TV industry was still entering new communities and laying cable. It was enough for a cable company to carry HBO, much less think about offering Internet or phone service. And the fractured local phone companies were trying to get on their feet, just as AT&T had planned. They posed little or no threat to Ma Bell.

Well, a lot can change in 16 years. Competition has sucked much of the profit out of the long distance business. AT&T, while still the biggest player in that business, knows it can't look to long distance for earnings growth. What AT&T needs are value-added (read here "high profit margin") services to offer. But most of the really good value-added services require a local connection to the customer. Suddenly, that old local phone service is looking pretty good.

AT&T could enter the local phone business again, and lord knows they have tried, but it is expensive and they are not trusted. They have bought companies and entered into deals, you name it. They have bought a big chunk of the cellular phone business. And now they have purchased TCI, the nation's largest cable TV company. It's all in an attempt to regain that last mile of wire to the home or business they threw away in 1983.The market is so much richer and so much more confused than it was back in 1983. Back then, there was local phone service, long distance phone service, rudimentary local cellular phone service, rudimentary online service, and cable TV. Today, all those formats overlap with the ability to offer phone service on cable, video service (of sorts) on the phone system, a national cellphone network that bypasses long distance, and then there is the whole new factors of the Internet and direct satellite broadcasting of both video and data.

There is a lot more than there used to be to sell over that last mile of wire.

But because AT&T can't just buy back those RBOCs it thought were valueless 16 years ago (even if AT&T could afford the price, it couldn't get regulatory approval), it has to sneak back into the local business through the guise of wireless, or cable, or a player to be named later. And at the heart of all these attempts to squirrel back into the local business is TCP/IP technology.

Crazy TCP/IP is in the news because it turns out to be the best way to handle all those different types of data at the same time. Want to mix voice, video, music, and AOL? Use TCP/IP. And with the Internet as an international backbone that goes places AT&T doesn't and costs AT&T next to nothing, IP telephony is suddenly an especially hot technology.Of course, there are tradeoffs. The early Internet phone products sucked. There was no gateway out of the Internet to reach phones that weren't connected to computers that were in turn connected to the Net. Even on computer-to-computer communication, the other guy had to be expecting your call. And the quality sucked, with lots of line noise, echoes and dropped words. They were free calls, sure, but free came at a stiff price.

That was before IP telephony became big business. Now there are IP telephony product programs at Cisco, Lucent and 3Com aimed at both offering better back-end support for IP telephones and better front-end systems, too. Aplio ships an Internet telephone that doesn't need a PC, for example. And lots of phone companies are building gateways from the Internet to the local phones systems of their countries. The Internet phone calls aren't free, but they can call regular phones.

Some of the technical improvements are impressive. 3Com just got a patent on a technique to help recover from dropped packets. Their algorithm includes some information in each packet about the packet before and the one after. That way, if a packet is dropped, it can be essentially recreated from the packets before and after it while sitting in a memory buffer waiting to be played to the listener. Such packet recovery makes no sense at all in the world of traditional TCP/IP, where a lost packet is just re-sent. But traditional TCP/IP doesn't put any premium on creating a continuous stream of what's essentially real-time data. In TCP/IP, the value was always on recreating the file exactly as it was no matter how long it took. That certainly doesn't describe a phone conversation with grandma.

So Internet telephony is improving at a dramatic pace. New products are appearing all the time. Now you can have your voicemail messages automatically sent to you as e-mail. There are literally Internet PBX's, which are a combination of an IP phone and a virtual public network. Big bucks are being invested on the idea that the IP network will carry every kind of business and home data in the near future.And that brings us back to AT&T and its lust for offering local service. Buying TCI brings tens of millions of cable TV homes within range of AT&T local phone and Internet service, thanks in part to @home. But cable TV has almost no penetration into business, where the really big phone bucks are spent. How does AT&T gain entry to millions of businesses, too?

Enter Novell, stage right.

Novell is busy adding IP telephony capabilities to Netware. Almost every business has an Internet connection today, and one important thing about Internet connections is the bigger they are, the cheaper they get on a per-bit basis. So why not trade that T-1 for a fractional T-3 and throw phone calls onto your LAN? And that's exactly what will happen.But wait, there's more! Depending on who you ask, Novell has from 50-80 million seats, each of which also has a telephone. Novell is working with AT&T, which will provide back-end network services. Now here's an interesting question: If you look at Novell not as a network software company, but as a way of buying access to 50-80 million seats, how much is the company worth? Cable and cellular phone companies are valued at thousands of dollars per user while software companies - even Microsoft, much less Novell - are valued at far less per user. What's wrong with this picture?

After languishing for five years and missing the biggest bull market in history, Novell shares have suddenly tripled in price. I think it is because of AT&T and IP telephony, as those in the know buy shares in advance of what they expect to be either an AT&T acquisition, or at least a revolution in how network software companies are valued.

Comments from the Tribe

Status: [CLOSED] read all comments (0)