External Modem Recomendation 3989

That might sound like a conspiracy, but (really) it isn't!

I can think of multiple scenarios that would have fit that description to a 't'. It is fairly common. One is that if digital switches had already been installed, the initial network topography was probably based on what had worked with old mechanical switches. That meant one "Central Office" switch per area, where "area" was determined by how far a local loop could be extended on typical outside cable plant. With different engineering examples, that might mean out to 5 miles, or even 10 miles. But the point is that a fairly large area would require multiple CO's, and there would be trunks between the CO's.

When major switch manufacturer's became aware of how that was being implemented, they came up with a "management tool" to "encourage" a more appropriate paradigm. (Switch manufacturer's, which basically means Lucent and Nortel, were *very* aware that a digital telephone switch is nothing but a big computer, while operating companies were totally unaware of that. Hence simply telling an operating company what would be best, wasn't viable. They had to make it economically best to do the right thing.)

The switch manufacturer's license the software that runs on those switches, and periodic upgrades are required. It is possible to miss some, but basically every 2-3 years there is no choice. By making each CO switch more expensive (well over a million dollars per upgrade) compared to software for a "remote unit", the manufacturers forced operating companies to re-engineer their network topology to have as few CO switches as possible and as many "remote units", controlled by a single CO, as it took to be functional.

What you describe sounds like the implementation of that topology. Either as they converted from analog to digital switching in the first place, or at a later date when they either re-engineered the entire topology, or when they implemented major expansions because of the increased number of lines required as business surged in the 1990's.

If they already had a significant investment in analog line interfaces for their existing digital CO switches, they would naturally try to make use of that investment. Likewise if the engineers had little or no previous experience with digital switching systems, they might pick essentially the same paradigm for ease of implementation. (Both of those last two were very common with the Bell operating company in Texas.)

That is a technically inferior method, and will not work with v.90 modems or with dsl lines. And it limits v.34 to 26kbps.

That does *not* mean that it provides "substandard" lines though! In fact, v.32bis is designed to provide a 14.4kbps connection on a loop that just meets minimum standards. Which is to say that if your line will support at least 14.4 kbps, it *definitely* meets the required standards!

Of course, that made good sense in 1985, and by 1995 was extremely annoying to hear. By 2005 it is abject stupidity on the part of any telco to irritate customers with such limitations!

Actually, almost any well grounded marketeer would have been able to sit them down and in a single four hour session explain, in detail with dollars and sense, exactly what was happening. Anyone who wanted to continue the "investigation" would be considered a bit slow... There is nothing to investigate; but it is a very touchy public relations problem.

Telcos all think that customers are necessarily "a bit slow". But that's because telcos are run by Pointy Haired Bosses... ;-)

Oh, it's far worse than you are imagining! AT&T just folded... because try as they might, they could *never* get control of the company away from people who grew up in the Long Lines department of a regulated monopoly. AT&T understood how to run a network for circuit switched message traffic. (Which is to say, voice calls.) They were *never* able to market anything.

Ten year ago every indicator they had was saying that "voice" was going away and "data" would be all that was left. Their only method to deal with that, was to guide the ship by looking off the stern to make sure the wake was straight! And that is how they ended up going aground.

Consider that when v.32bis modem protocols were originally designed everyone who knew about The Internet and understood where the future of computer networking was headed, knew that 1) standard telephone lines were able to deliver multi kilobyte data rates between the CO and the customer, and 2) digital switches and long distance were all limited to 64kbps per connection. When the design for ISDN (in the mid-1980's) came out with the ability to provide customers with one or two connections at 64kbps per connection... telephone company (Pointy Haired Bosses) management simply could *not* see a market for it! "Who is going to pay for it?" was the question about implementation. As a result virtually no telcos implemented ISDN in any reasonable way.

So, instead of the telecom industry providing for customer needs and grabbing the entire market, modem companies first (with v.34 and then v.90 protocols) and then even cable TV companies, jumped on the bandwagon to supply Internet bandwidth. They made millions, and the telcos sat there until well into the mid 1990's without a clue... and they missed being able to control a multi billion dollar market that they could have locked down right from day one.

Satellite has disadvantages too. Shared bandwidth and high latency, just for starters.

At 15 miles from the telco, you necessarily *have* to be connected to a "remote unit". The catch is in two questions. One is how far are you, in cable miles, from that remote? It would probably be a little hut, maybe the size of a pickup truck, but could be either larger or smaller. The second question (and much harder to get answers for) is what type of facilities do they have between the remote and the CO?

If they do have some purely digital facilities, and if you are less than say 3 miles from the remote, it *is* possible to get better connect rates. That is why the question about does *anyone* at all get better connections? If *nobody* does, it may just plain not be available. If somebody does, the trick is to find a way to get your line provisioned on the facility that provides better connections.

One example of a way, which is not without expense but in some places has been reasonable enough to do it, is to order up an ISDN line, and when it works, cancel your other line. Then wait a month (long enough for them to re-use the facilities for your original line), and cancel the ISDN by going back to a regular phone line. They most likely will use the digital facilities for the ISDN line to provide your new service... and you'd then be able to run a v.90 modem on it.

Of course, if ISDN installation costs are exorbitant, that is too risky to try unless you know it will work. The trick then becomes one of finding a friendly telco engineer who will clue you in on *exactly* what you have and what is available. Such a person might even just switch you from one to the other to be a good guy. Or you may have to put 2 and 2 together from what you hear from 4 different people to get an idea what will work.

Or, that kind of politics might be just boring as Hell. You can get a good modem and live with 26kbps for a few more years.

Now, just to make you feel bad... I worked for a long distance carrier for more than 3 decades, and was intimately involved in how this whole scenario took shape here in Alaska. I was not able to use v.90 until about 3 years ago. But, whoopee, by early this winter it appears that I may actually have some kind of broadband service! (Long after I've retired, and provided by the "other" long distance company! What a hoot.)

--