Are you holding your breath waiting for this next big thing? Better have a medical technician standing by.
Ted Prince
Are you getting as sick as I am of the daily announcements about ATM (asynchronous transfer mode) products? Or even worse, ATM-enabled products, whatever they might be. I mean, is anyone except a few universities and corporate pioneers actually using this stuff now, for real-world applications, all the time?
Doesn't this look like deja vu all over again? It's just like when ISDN was announced in the early 1980s. When everyone was talking breathlessly about the integration of the computer and the telephone. When IBM bought--and then divested--Rolm.
Lest we believe that ATM is right around the corner, think first of the problems of moving to ATM, or fo
r that matter to any other all-encompassing megatechnology. Interoperability issues are huge (we still have major interoperability issues with ISDN after 10 years of a supposedly established standard).
The cost of the ATM stuff is disproportionately high, and so are the connection and transport prices (all of which have been major issues stalling ISDN). Skills in installing, using, maintaining, and debugging ATM are nonexistent in the marketplace at large and will be for many years (just like ISDN again, even in the telephone companies themselves). Add to that the complexity of ATM-only networks, let alone heterogeneous networks, including everything from X.25 to video-on-demand.
In any case, why do normal people need ATM anyway? Isn't there more than enough of a choice out there? Frame relay, fast Ethernet, SMDS (Switched Multimegabit Data Service), not to mention ISDN. Unless you're transferring the entire MGM film library to every branch office in the world every day, aren't these older techn
ologies going to cover just about everything you expect to do over the next few years? Is ATM a solution in search of a problem?
Even if ATM finds that killer problem this week, how long is it going to take to become a well-established standard? The telephone companies haven't done that with ISDN in 10 years. Let's be optimistic and say the pace of change has doubled--it will still be five years at least before ATM is established.
So why do we need ATM now? Is it the new corporate MIS status symbol to show that we're with it? Is ATM just the latest shot in the eternal class war, this time played out in that new frontier called cyberspace?
In the real world, the success of a technology is not a function of its engineering elegance. Witness the Concorde 20 or so years on. Ultimately, the test is whether or not the technology is appropriate--something we technofreaks tend to forget. ATM is clever. ATM is elegant. But is it appropriate? Right now it isn't.
What with the Internet, on-li
ne services, and lots of PC-based communications methods and standards, cyberspace is fast becoming a protocol soup, where no one communications protocol dominates and newcomers are even less likely to do so. No one delivery protocol is going to take over the world of communications, nor should we want it to. We already have a delicious protocol soup with lots of ingredients that please many different palates. Some are gourmet, some are more basic. ATM is going to be just another, albeit gourmet, ingredient of the protocol soup, something to be enjoyed and savored by the technoliterati and the corporate titans while the peasants get by with the meat and potatoes of frame relay and ISDN.
The protocol soup that is communications today and tomorrow reflects a wide variety of demands and needs that the soup is by and large meeting. New ingredients will also be added: faster and longer Ethernets, ATM Lite, Son of Frame Relay, SMDS for TV, broad-ISDN, and so on. Each will meet the different needs of differen
t users.
So you shouldn't just blindly jump on the ATM bandwagon. There's more to communications than fashion. Figure where you stand in the spectrum of needs and plan accordingly. Take into account the rich variety of existing methods that fit many needs very well and much more cheaply than ATM will. Leave the hard ATM stuff to others. Wait until it forms a real part of the communications environment and isn't just another brilliant but expensive technology waiting for its time in the sun.
photo_link (40 Kbytes)
Ted Prince is president of Perth Ventures and publisher of
Technology Fundamentalist
, a newsletter that covers "winners and losers in information technology." You can reach him on CompuServe at 74073,1236 or on the Internet or BIX at
editors@bix.com
.
ATM = After the Millennium
