DSL or Digital Subscriber Line is simultaneously the most popular and most criticized broadband technology in use today. It's popular simply because it's cheap, at least compared to its predecessor, T-1. The criticism side of the equation gets a bit more complicated. One interesting quirk is that everybody, myself included, erroneously calls it a DSL line, in spite of the fact the word "line" is already built into the acronym. (Soon, I'll be entering a twelve-step program for abusers of TLAs - Three Letter Acronyms.)

The first thing that you find out when you research DSL is that there is almost too much info. (Begin at www.dslreports.com and go on from there.) There is a whole bunch of different subtype within this technology, each with its own acronym, and lots of picky technical details. Luckily, only two subtypes are really common, SDSL (Symmetric) and ADSL (Asymmetric).

The biggest difference between them is in the way speed is given. SDSL has the same speed upstream and downstream while ADSL can be different, with the upstream speed slower, usually by a lot. While it's not usually sold that way, there is no law that says you can't pay the extra price to match speeds on ADSL.

The other difference in the two is that of target marketing, based on their underlying characteristics. SDSL is aimed directly at business, and is typically bundled with all of the extra features that a business needs. ADSL is aimed at individual users, and to a limited extent, at very small or home-based businesses. The business consumer has to be wary, however, as ads for ADSL sometimes do everything they can legally get away with to obscure the fact that they are for ADSL services.

While SDSL speeds are not absolutely guaranteed, they can be depended upon to stay at very close to the maximum advertised rate. So far, the ones that I have tested have been at 95% or higher. ADSL, on the other hand is cloaked in terms like "as fast as" or other such weasel words. During "peak" times, however that is defined, you might even get less throughput than you'd get with a modem. Even during "non-peak" times, maybe you'll get the rated speed, and maybe a lot less. Also, the contracts usually let the provider throttle back at will. That's why businesses usually don't bother with ADSL.

You may have noticed that I have, so far, neglected technical issues, especially the specifics of speed. That's because things get a bit muddled here. I'm going to play it safe (and short) by using mostly concepts, not specifics. Feel free to look up the details on your own.

First, DSL speed is dependent upon the distance from the phone company office to your site. The farther away you are, the slower the speed, with SDSL having shorter limits (usually quoted as 12,000 to 16,000 feet max) than ADSL (usually quoted as 16,000 to 18,000 max) feet. Inserting a phone company emulation device in between the real phone company and the end user can alleviate this. Phone companies in areas with high market growth possibilities are doing this. Most SDSL and ADSL lines peak out at 1.5Mbps (millions of bits per second) for short distances, but I've seen 2.3Mbps and even higher advertised. Newer types off DSL are faster, but are still quite rare. Note that the distance to your site listed in the phone company's computer might not be accurate.

Another way around the distance limitations is the DSL variant-of-last-resort (my term) IDSL (ISDN DSL, where ISDN is Integrated Services Digital Network). It can reach out beyond 21,000 feet, but can only run at 144Mbps both ways, and costs just as much as SDSL.

The distance limits are somewhat esoteric because they combine actual physical limits and the limits imposed by providers local phone companies due to either reliability worries (longer is less reliable) or marketing (hence, "usually quoted as"). Cheap DSL service can potentially replace expensive T-1 service, thereby putting a large dent in phone company profits.

Other physical limitations can also prevent DSL service. Specialized signal repeating or filtering devices placed between your site and the phone company will disqualify a given line. Segments of fiber mixed in with the copper cable can also prevent DSL, since the specification is for copper cable, but a few companies have found ways around that. Since some areas have run out of end-to-end copper lines, and many phone companies and providers can't seem to coordinate in cannibalizing existing lines, this could keep you from getting DSL. Additionally, the selected cable might not test out as being of a high enough quality to use for DSL.

The actual task of coordinating the installation of DSL among the phone company, the provider, and, where necessary, a third-party intermediary firm, can cause delays of many weeks in getting a DSL line put in. There is also the added risk of misunderstandings or disputes causing further delays.

Now here is the fun part! DSL providers and their third-party intermediary partners are dropping like flies. If one of them goes under while you are connected, you could lose your line. In many, though not all cases, phone companies and some larger providers are cherry-picking the most profitable areas from these failing companies and either reconnecting customers or keeping them from getting disconnected in the first place. If you are not in one of these areas, or the company goes under too fast, you are out of luck. Phone companies, of course, seem to be doing their best to encourage the demise of these perceived competitors.

In spite of all of this nonsense (and even more I don't have room for), I still recommend DSL as the first choice in broadband for my clients. It may not be as reliable or as well supported as T-1, but it's cheap enough to be attractive despite its limitations. All I ask is that they monitor the news about their provider and any intermediary (If I sold them the line, I do that.) and keep a filled-out application form for an alternative provider handy, just in case.