Index Home About Blog
Newsgroups: comp.dcom.isdn,
Subject: Re: DMS-100 BRI upgrade sked for 4q98 - 8 terminals on a line
From: (Floyd Davidson)
Date: 18 Nov 97 10:05:29 GMT

In article <>,
David Lesher <> wrote:

>  (Laurence V. Marks) writes:
>>I only have this on hearsay, but I'm told (Fred G., I think) that the reason
>>that the LECs prefer Nortel is that the software is easy to use.
>They were supposed to be far cheaper to buy, much cheaper to run
>[power req and HVAC, and floor space], cheaper to upgrade [re: code
>update price], and less maint hours to tend.
>I've heard they fell short in several areas, but that's skuttlebutt.

Hidden in that conjecture is an interesting paradox which
developed over time with the development of DMS switches.
Originally, in the late 1970's and on through to the late
1980's, Northern Telecom, Inc.'s number one selling point was a
marketing claim of "no maintenance required", meaning a vast
reduction in manning allowed by moving to the DMS platform.
That is exceedingly significant because NTI managed to latch on
to about 40% of the market share during that time frame, but it
lead directly to a conflict between marketing and development.

During the early and mid 1980's NTI's marketing stance was that
a DMS switch required virtually no maintenance, therefore they
would allow NOTHING which appeared to contradict that position.
Software to facilitate administration of maintenance operations
was a subject they would not mention!  Even interfacing the
switching system to an external computer system for those
purposes was verboten!  That marketing strategy would not allow
software tools to be developed for use by an on site maintenance
crew that didn't exist in their view.

As a result most DMS systems were installed and managed under
the concept that administrative computing services were not
required and computer unaware upper level management (and
usually lower level management too) in most cases simply did not
understand how useful it would have been because NTI said it
wasn't needed.  (At that time most telco management didn't
even realize that a digital switch was in fact a computer,
so the connection between administrative computer services
and a switch didn't exist for them.)

But by the late 1980's the digital switching market had been
divided up, and a shift from finding new customers towards
finding ways to keep old ones and expand their "investment
opportunities" became more important.  At that point NTI began
designing software tools to manage and administer maintenance.
It took half a dozen years to mature and it wasn't until the
well into the 1990's that such software tools became ubiquitous.

Relatively speaking, all along the way the software for the
DMS switches has been very nice, from a maintenance point of
view.  The problems of getting management to understand that
software is useful as a tool, however, can be great!  I have
heard of many offices where either disk space was write
protected for maintenance logins, or where writing a command
macro and saving it to a disk was consider a firing offense!
Can you imagine having a computer system the size of a DMS-100
and being told that using the command line interpeter for
scripting would get you a pink slip!


Floyd L. Davidson   <>   Salcha, Alaska

From: Floyd Davidson <>
Subject: Re: Any step-by-step offices in the US or Canda left?
Date: 26 Aug 2000 02:39:45 -0800

Crossbox <> wrote:
> (Lisa nor Jeff) wrote:
>> Are there any electro-mechanical switching offices left in
>> service in the US or Canada?
>> One would think in a remote area there might be some, but
>> those places may have been the first to switch over since
>> maintenance costs drop dramatically.

>Second, your logic about the order of switchover is incorrect.
>The first 1Aess came online in a fairly urban area of Illinois,
>if I remember right.  The more dense the population, the more
>need for a switch upgrade, and the technology always trickles
>down.  That's how it works with most everything.

Trial installations might well have been in settings where an
initial installation could be well monitored, but that is not
what would be meant by "those *places* may have been the first".

Indeed, maintenance costs were what drove the early marketing of
digital switching systems.  That was true in particular for
Northern Telecom, which began with zero market share in the
(essentially brand new) US market.  They eventually acquired
something like a 40% market share, which was mainly accomplished
by touting their digital systems as "maintenance free".

That led to two concurrent events, which were more or less at
odds with each other, relating to Northern Telecom switching

First, areas with high labor costs naturally were most attracted
to that marketing approach.  Here is Alaska, where "windshield
wiper time" can be exceedingly expensive (I have sites that cost
more than $1000 just to put a technician on location), virtually
all switching systems outside of major urban centers went
digital by about 1985.  All inter-switch (toll and autovon both)
switching was digital in the early 1980's.  At the same time,
country wide roughly 33% of all lines were being switched via
digital systems, and just as in Alaska, virtually the entire
long distance network went digital almost overnight...  except
that AT&T used 4E switches while in Alaska it was DMS-200's.
Locations where labor costs were low retained non-digital
switching systems.

The second effect was that NTI engineering was not allowed to
introduce features clearly targeted at a maintenance force!  In
theory, there was no maintenance force, so how could a sales
person pitch features for one without giving away that fact that
they all knew a maintenance free switching system actually did
need a full time maintenance staff?  That meant discussing
enhanced maintenance and administration potential, for example
by using computers or computer networks, with NTI was a total
dead end!  They would not even admit you could connect a
computer to a DMS switch, much less that there would be any
benefit from it.  And of course later Scott Adam's Dilbert
cartoons were very descriptive of telecom management mentality
at the time, and very few switching systems ended up with
computerized alarm monitoring, log analysis, or testing
facilities simply because NTI would not discuss it and telecom
management didn't have a clue on their own.  It was frustrating
as hell for maintenance people who were also computer aware,
because while NTI would not admit it could be done, it was also
very clear that *they* did it!  Switch log formats were clearly
designed for computer analysis, and anyone with programming
experience could imagine in an instant dozens of ways to make
use of a general purpose computer tied to the switch.  But in
most companies not only was that not allowed, but some would
even fire technicians for using the built-in shell scripting
facilities (which were fairly good).

That marketing approach was maintained right up to about 1989,
and the effects of it on telco management perceptions were still
in evidence for several years afterwards even though NTI changed
directions and began to market added value to an existing stable
customer base which they no longer expected would grow.  Today
of course NTI sells such computers as a peripheral devices that
directly access the DMS bus.

Floyd L. Davidson                
Ukpeagvik (Barrow, Alaska)

Index Home About Blog