The Great Failure of Residential ISDN in the US Market

6 Replies

Integrated Services Digital Network, or ISDN, was going to be the next step, the incremental movement beyond modems.

Because incremental was the norm.  We were used to that.  We had a whole industry that went from 300bps and 1200bps hardware to 2400bps then to 9600bps then to 14,400bps then to 28,800bps then to 33,600bps, finally capping at 56,000bps some 20 years down the road.

And modems were capped at 56Kbps due to the bandwidth of phone lines in the US.  And even hitting that speed was a bit of a pipe dream.  Modems commonly negotiated down to a lower speed based on line conditions once 28.8Kbps became the standard.  The firmware on a typical 56Kbps modem was configured to return CONNECT 56000 on any v.34 connection, regardless of the speed it actually settled on… and unless you had no problems in your home wiring and were connecting to an ISP on the same phone switch in the same central office, that speed was not going to be 56K.

If you worked on modems you knew that was the truth, that what you could do in the lab on a line simulator could be miles from what was going on in the real world.

But ISDN was going to solve all of that.  ISDN was all digital… that “D” is for digital, just like I said in the first sentence of this post… and so if it said it was going to deliver a 64Kbps connection, you got a 64Kbps connection.  And, you didn’t have to settle for that incremental upgrade… it is hard to sell going from 56Kbps to 64Kbps even if you know that the 56Kbps value is a lie and not at all comparable to the low latency, solid 64Kbps connection that ISDN could deliver… because with each residential ISDN line you got TWO 64Kbps connections, called B channels, for bearer channels, which could be paired up for 128Kbps AND you got another 16Kbps D channel for signalling and other side data needs.

And if a call came in, ISDN would tear down one of the B channels, pass through your voice call, and you could continue on doing whatever online at 64Kbps while talking, then when the call was over you could get back to full speed.

The promise of all of that, plus the D channel, which could potentially pop up news or sports scores or whatever… it was a 16Kbps secondary channel… opened up a fantasy world of possibilities.

Then the phone companies got involved and fucked the whole thing up.

Okay, the fantasies were a bit grandiose at times and today, where I have a 400Mbps line pinned up in my home 24/7 for always online internet (Are you old enough to remember when people used to brag about their company having a T1 line for data, which was 1.544Mbps?) the idea of even a 128Kbps connection seems piddling.  At the right moment in history though, that speed would have been something.

If the phone companies could have gotten on board with large scale residential ISDN around, say, 1990… or even 1993… it would have been a boon for the end users and would have sunk the modem market early.

The truth is though, while there were pockets of experimental deployments, the demand wasn’t there and the phone company wasn’t interested in investing without that… and when the demand was there… well, the phone company got greedy.

When we were facing the great dial tone drought in Silicon Valley in the mid-90s, the phone companies were suddenly regretting the whole flat rate dialing plan idea that they had come up with in the late 60s.  That plan was great when phone calls rarely lasted more than a few minutes and their infrastructure didn’t need to be scaled up even as they were hooking up more residential phone lines.  It was a reliable, predictable revenue stream that made both them and the customer happy.

And then the internet showed up and a bunch of people started staying online all evening or even all night, reading email, checking for email, downloading stuff, checking for email again, looking at porn, checking for email once more just in case, playing a video game or two, and arguing with people on Usenet… and the infrastructure wasn’t up to the task and the customers were not happy and the government boards that oversaw the phone company’s happy local monopoly started asking awkward questions, and soon then had to actually open up their pocketbook and start upgrading the network they had been largely neglecting since the break up of AT&T.

So the phone companies swore they wouldn’t get suckered again with a flat rate dialing plan for any new service, and the next new service on the agenda was ISDN.  That was going to be a metered connection, pay by the minute, make those internet addicts pay pricing plan.

Flat rate, all you can eat at 56Kbps… even if that ends up being 24Kbps for a lot of people… suddenly looks very attractive when compared to 128Kbps when the phone company is looking to bill you in six second increments or some such.

This played out differently in different markets, and I can only speak to how PacBell operated, but the net result was that demand for residential ISDN was suppressed.

Here in the domain of PacBell the Public Utility Commission viewed the whole metered connection thing as the cash grab it was and reigned in PacBell, making them offer a some of the connection time at a flat rate, but allowing them to have a data cap, after which they could charge more.

PacBell had to go along, but they did so by embracing ISDN installation the way one might embrace a week old halibut found laying on the street on a hot summer day.

You could get an analog phone line installed the same week, but if you called up and asked about ISDN you might get no answer at all because they weren’t sure where it was available or who might install it.  When you could get an appointment, staff were untrained and admitted as much… they were not happy to have been assigned a task they hadn’t been trained for… so your chances of getting a live line on the first try was dubious.

I happened to get lucky.  When the installer came out for my appointment he had already done a couple and knew what to do.  This made me an outlier at the office where tales of failure were the norm and it took two or three tries for some people before they had a working line.

At my place the junction box was close by and even though the installer wasn’t sure what the SPID was for my connection… that process was not yet codified or some such… I had on me a patented… seriously, PATENTED… document of instructions on how to calculate the SPID based on a set of known parameters and standard behaviors of various phone companies, which turned out to work.  So I was up and running on ISDN on the first try.  I had a Motorola terminal adapter… it isn’t a modem because there is no modulation by default, at least until you attach an analog phone to it… because the Boogie Board was still a work in progress.  But we wanted to have knowledge of the process of getting residential ISDN set up.  We had been through the whole caller ID thing without caller ID being available and were still smarting a bit from that.

It was a unit like this… maybe the Bitsurfer… it was a long time ago

And the hardware was another thing.  Unlike a modem, which you can go buy at the store and setup immediately with your phone line, the ISDN terminal adapter is of zero value until you have your ISDN line wired up and configured correctly.  And, thus, ISDN was also not an impulse purchase, and I cannot emphasize how much of tech progress is driven by impulse purchases by males with more hope than sense.

But when it all came together, it was sweet.  In the age of modems, ISDN was head and shoulders above analog dial-up.  My ISP wanted to charge me extra for 128Kbps connections, but let me ride at the same price as normal dial-up for a single channel 64Kbps connection, so I lived with that.  Playing early online games like Delta Force or Starsiege: Tribes on even that single 64Kbps line gave me a serious advantage.  Playing Diablo II online with friends I could see their locations getting updated more slowly compare to my own.  And when EverQuest came out, oh man.  I used to zone so much faster than other people.

It came at a price though.  When the tech wired up my ISDN connection, he punched down my pair of wires at the far end of the box, by themselves, to indicate that they were a different service.

However, that did not translate to other techs showing up.  They would show up, see a pair of wires hanging out by themselves at the far end of the junction box and think, “Cool, an unused connection already setup for me.  I’ll just wire up this order I have with those and be done in record time.”

Then I would get home from work and the ISDN would be down and I would have to call PacBell who would get to it in a few days.  It used to be that I would be driving home and I would see a PacBell service truck on our street or driving away as I pulled up and I would know that the ISDN would be down… and I was right every single time.  And about half the time the tech who came out for my service call wouldn’t know what the hell was wrong or what ISDN was, so unless I stayed home for the service call there was a good chance I would have to call back again and complain that they didn’t fix the problem.

When it was up, it was sweet.  When it was down… well, a down line is of no use at all.

And I know that this is my own experience in the SF Bay Area in the back half of the 1990s… though you might be excused for believing somehow that Silicon Valley would be on top of this stuff, being the tech capital of whatever.  Other places in the US fare differently, and I know that in Europe there were countries with very high adoption rates.

The problem in the US was that once the phone companies started to get their act together on the whole ISDN thing… other technologies were already becoming available.  Sure, I zoned fast in EverQuest, but not as fast as a guy we used to group with who had this thing he called a “cable modem” from his local cable TV provider.  My first brush with what we now think of as broadband.

Which brings us to Big Island and our ISDN ambitions.  We were some keen early adopters… relatively speaking for the area… of the technology, but even as we were getting our first product out the door, the YoYo I wrote about previously, PacBell was announcing ADSL, a flat rate, always on digital data connection that could be run on the same wires… simultaneously… as your analog phone line.

I seem to recall the PacBell announcement was the same day as the YoYo launch press release, but I could be wrong on that.  It coincided with some event at Big Island and it was a turd in the punch bowl of our dreams.

I kept my ISDN line for quite some time, until my wife and I bought our first house.  The new place was too far from the local junction box to run either ISDN or ADSL and I had to return to analog modem dial-up.  And if the copper pair is too far to run those digital services, then there is no way in hell I was going to ever get anything close to a 56Kbps connection.  It would be a while and a couple of tries before something like real broadband was a thing for us.

Big Island’s ISDN card, the Boogie Board, and all the wonderful things it was going to be able to do… as I said, everything the YoYo did on steroids along with a lot more… fell by the wayside.  The VCs who had been financing the company cut way back.  They shifted some of our tech… which they owned… to another startup they were funding.  We were not selling enough YoYos to pay anybody, so if we couldn’t hold out and work for free we were going to be looking for another job.

And, our there somewhere, somebody was annoyed enough by the collapse of the company to write to the US copyright office to ask that commercial software source code be put in escrow so that when a company goes away there will be some way to get the software updated.  I am kind of with him on the Claris Emailer front, which happened to be the software package I used to do all of the email support for Big Island.

Putting software in escrow… that is an enterprise software thing, and we’re not there yet.  Before we get there I had to go find another job.

The Telephone tales so far:

6 thoughts on “The Great Failure of Residential ISDN in the US Market”

  1. PCRedbeard

    Are you old enough to remember when people used to brag about their company having a T1 line for data, which was 1.544Mbps?

    Yes, I do. Having a T1 line in the mid-late 90s sure was something else. In the pre-YouTube days, streaming bandwidth wasn’t that bad, so you could get a helluva lot done at work on a T1 line, even at the mid-sized company I worked for back then where during March Madness you’d have about 500+ people updating scores (or occasionally streaming a radio broadcast) at a time.

    Like

    Reply ↓
    1. Wilhelm Arcturus Post author

      I think PointCast was the first app that really taxed office connections and got IT to put its foot down on what you could and could not do at work.

      For about 18 months, from early 1993 until we moved to a new office in mid-1994, Global Village’s connection to the internet was a single 56K leased line. That was enough for a couple hundred people to do email, Usenet, and some early web surfing.

      Like

  2. Nogamara

    Fascinating read. I’m not sure if Germany was the actual poster child of ISDN but it sure felt that way. I suppose I only experienced the tail end of this (we got ISDN in ’98) but it felt like many people had this as their first means of getting online, but just as many switched from 56k. And I don’t remember 56k being bad (or people not complaining, but it’s been 25 years…), so I guess cabling was better.

    Anyway, even with dialup we mostly had minute prices, so no big changes here, althouzgh I remember some time-limited offers of “you only pay once for dialing in, not per time” – so that was night time stuff.

    Like

    Reply ↓
  3. flosch

    > And if a call came in, ISDN would tear down one of the B channels, pass through your voice call, and you could continue on doing whatever online at 64Kbps while talking, then when the call was over you could get back to full speed.

    Wow. I used ISDN for several years while I was still living at my parents, and then I think the very first time after I moved out… ISDN was actually a big thing in Germany, to the point where it very much delayed DSL or any other faster connection in the early 2000s, but that’s a different story… but I had never heard of this functionality of automatically disconnecting one channel, then later re-bonding them.

    I wonder whether that was another thing that the telco screwed up and didn’t let us use? It’s interesting, because I’m pretty sure (though my memory is getting hazy after that many years) flatrate Internet access (or any kind of phone call really, no local flatrates in Germany) only arrived way after ISDN did, so the few times I used bonded channels, I paid double.

    Like

    Reply ↓
  4. Archey

    I admit, I was one of those bragging about the T1 line my employers had back then.

    This is an insightful post though if only to understand why ISDN never became a thing. By the end of the modem age, I was finishing college and moving from my small Louisiana town to a nearby large Texas town. I hoped upon moving to a big city to get faster internet, but my first apartment complex had some deal with an ISP that precluded ISDN, ADSL, or anything else. I moved to a 40x larger city but still had the same dial up speeds until cable internet became ubiquitous.

    I still think of the acronym as standing for “It Still Does Nothing” all these years later.

    Like

    Reply ↓
    1. Wilhelm Arcturus Post author

      The funny thing is that ISDN is all over the place now, but only for trunk lines.

      One of the things that kept US Telcos from adopting ISDN is that with a T1 line you have to give up one of the 24 voice channels for signalling. In Europe, where their later E1 line protocol had 30 voice channels and 2 signalling channels… because ISDN was already in the works… there was no loss in capacity with ISDN while in the US you lost 1 in 24 voice lines.

      The superior signalling eventually won out in the US in the early 2000s because big customers, like call centers, were demanding the features it offered.

      There were no huge obstacles to the US adopting E1 back when it first showed up, but it would have required time and money and the cooperation of a lot of little regional phone companies, if it had been done early enough. But the two protocols use incompatible voice encoding (Mu-law vs. A-law) which gets handled on international calls through conversion at a connection point, but adopting it in the US would require every voice prompt on every horrible old IVR to at least be re-saved in the new format.

      Though, in the end, all of that is kind of going away as IP replaces those old telephony-centric protocols.

      Like

Voice your opinion... but be nice about it...