Why DSL Still Sucks

Practical explanations and compelling conspiracy theories.

Journalism doesn’t get more anecdotal than this, but it seems like everyone I know with a DSL connection to the Internet has bitched about it. At length. To me. Can’t get connected. Can’t stay connected. Got connected and then, mysteriously, lost the connection for weeks and weeks, and no one could get them back on.

Just how hard can it be to put people in the fast lane and keep them there, anyway?

A flurry of acronyms must accompany any discussion of telecommunication issues, so please bear with me. DSL stands for digital subscriber line, the technology that connects you to the Internet through your conventional telephone line. This dependency on conventional telephone lines is where the connectivity problems begin for many DSL customers. As the phone lines leave your house or office, they join an infrastructure consisting of wires and telecommunications equipment of varying quality, types, and ages, all owned and operated by your local exchange carrier (LEC, typically the local Baby Bell—such as Verizon, Qwest, and SBC).

Your LEC also owns the destination of all these wires: the centraloffice. The CO is just a big building where telephone wires from all the homes and offices in your service area meet. In the old days, the CO was where an operator sat at a switchboard and connected your wire with that of the other party and, eventually, long-distance lines. Over time, the switchboard became automated.

The 1996 Telecommunications Act required your LEC to connect your phone line to competitors’ services.Suddenly DSL providers—both independent operators and Baby Bells—sprang up in droves. Note that the Baby Bells could have introduced DSL at any time but chose not to until there was competition.

Into your house these providers installed DSL modems, which turn Internet traffic into high-frequency signals that can be transmitted across normal phone lines alongside lower-frequency voice traffic. Into your central office, they placed boxes called DSLAMs (DSL access multiplexors), which decode those signals back into Internet traffic and then send them on to a high-speed Internet connection. Simple enough.

Regulating Boxes and Wires
Actually, not that simple. The same regulations that made them open their CO to the competition allowed the LECs to establish inscrutable rules to protect the security of the building and the sanctity of the machines and cables therein.

The scattering of responsibility and lack of reliable service from the LECs make it tough for a DSL provider to debug a bad connection. Is it in the phone line? A bad switch at the CO? A broken DSLAM? Or is it a funky user setting on the customer’s PC? Or the wires connecting the PC to the DSLAM? Only the LEC can fix problems with the telephone lines carrying your DSL service. But they aren’t going to hurry, and they have no incentive to do it right the first time if it’s going to benefit their competitors.

The Baby Bells own the CO, the wires, and their own high-speed Internet connections. They are uniquely well-placed to provide their own DSL service. Robert X. Cringely argued a year ago that through a combination of poor service and price gouging, the Baby Bells were attempting to corner the market on DSL for themselves. Whether he was right or not, many of the high-profile independent DSL players (Northpoint, Rhythms) have gone bankrupt, and the Baby Bells are now the largest DSL providers by far.

Conspiracy Theories So, let’s say your DSL provider is a Baby Bell. Everything should work great now, right? Wrong! Things still go wrong. Horribly wrong. Why? Pick your theory.

Theory 1: Protect the Cash Cow
Baby Bells currently make good money installing high-speed T1 lines to connect (mostly) business customers to the Internet. T1 service is about five times as expensive as DSL service. Why should the Baby Bells cannibalize their own business by making DSL attractive to their T1 customer base? As noted, they were in no hurry to commercialize DSL before competitors arrived—and they still aren’t.

Theory 2: Dollars and Cents
DSL is expensive to install but even more expensive to support, so your DSL provider does everything it can to minimize the cost of fixing your connection when it goes down. It’s much cheaper to have a service representative walk you through reinstalling the DSL software over the phone than to send a truck out to inspect the wires. Every time the guy on the DSL helpdesk stalls your request, he saves his company money.

Theory 3: Emerging Technology
The roll-out of any new communications system is both technically difficult and expensive. For example, it wasn’t until the early ‘90s that cable TV approached the reliability of voice telephone service. Cable TV pioneers from even the ‘80s remember their screens turning to snow periodically, and cellular telephone pioneers can recall many more dropped calls in the early ‘90s than today. As any new communications system expands in its service area and acquires a critical mass of subscribers, it tends to become more reliable—in part because the company figures out how to stamp out all the bugs and because a large subscriber base gives it all the resources it needs to keep the whole system purring. If this theory is correct, the good news is that we can expect uniformly excellent DSL service, but the bad news is it may take the better part of a decade (groan).

Until That Day Arrives …
If your DSL service is giving you fits, see DSLReports’ trouble-shooting FAQ. The more you know about your hardware and software configurations, the less time you’ll spend dinking around with customer service on the phone, and the sooner a technician will arrive to solve your problems. Also, find a DSL buddy who has already been to war with the local providers. She might save you a lot of time and trauma. A final hint from a friend of mine who went through the DSL wringer: When you call to report a service disruption, ask first thing if there is a known service problem in your area.

The Competition
Are cable modems the answer to DSL woes? Well, it’s not exactly out of the frying pan and into the fire, but as a cable modem customer, I can tell you that cable is not the epitome of reliability. However, the technology is simpler, the lines on which the service runs are newer and therefore less buggy, and cable companies—like the Baby Bells—can’t pass the buck if the service doesn’t work. The consumer market seems to prefer cable: Jupiter Media Metrix reports that 6.6 million households connect to the Internet with cable versus DSL’s 2.8 million. To explore the advantages and disadvantages of each service, see thisDSLReports page.

Voodoo Economics
Like voodoo on a corpse, telecom regulations gave DSL life. But like a zombie, DSL currently walks the earth more dead than alive. Whether it, rather than cable or yet a different technology, truly becomes the mass-market broadband solution depends on the willingness of telecom providers to invest in their infrastructure. It also depends on your willingness to demand good service from your provider. Keep them honest.