[inforoots] Network infrastructure [2 of 2]

[Carl Baltrunas]

Hi John,

Interesting figures on actual cost.

What is missing is the development and maintenance costs of building 
and running the infrastructure.  Without that development of 
technology, the training of engineers on installation, maintenance, and 
refurbishing, and the manufacturing costs of the wires, switchers, 
terminals (aka phones), and
other parts, that incremental cost of just one phone call would not be 
possible.

I worked at Tymshare/Tymnet and it's associated companies from 
1980-2004.  Prior to the internet, it was the world's largest privately 
owned data network.  (I was told that the German Bundestpost was 
larger, but it was a public/state-owned company).  There were real 
costs involved with connecting each customer from  manufacturing the 
equipment, having lines run, having telco lines leased, provisioning 
and planning for expected bandwidth from a particular customer 
(sometimes after the fact, when they found specific circuits were 
overloaded), maintenance personnel to install and maintain equipment, 
even if it lay idle part or all of a day, week or month.  Interestingly 
enough, one of the services we offered to AT&T was called COEES 
(Central Office Equipment Evaluation System (or Service)) where they 
used our timesharing service to plan and order their own hardware for 
the telephone infrastructure, and each and every local telephone 
company continued to use after the Bell breakup.  That service was 
mandated by judge Green to remain available until the end of 1988 or 
1989, and I went onsite to AT&T HQ in NJ in '89 to assist them on 
migrating part of that service to a different platform on their own 
equipment which was co-located at one of our data centers at the time.

Yes, I agree, that once the equipment was connected and running, the 
cost of a call (network data call), was negligible, just as it was for 
the phone company.  However, they need to recoup the actual costs 
somewhere, and I'm not sure if any of these companies know how much the 
true cost is or was.  The simple answer is that any profit they made 
was over and above the true cost, but some costs are amortized and the 
real life of a component was likely NOT taken into account, but was 
taken at a particular cost for accounting (tax) purposes.  Thus, when 
an asset had a book value of $0.00, they made more of a profit, since 
they typically didn't lower the prices just because some assets were 
completely depreciated.  I'd easily say that the real costs were for 
personnel, and not for infrastructure resources used (except when 
lines/circuits were overloaded).

I recall a $20 fee for a magnetic tape mount, because the operations 
center decided that that was how much they needed to pay a full-time 
operator for the number of requested tape mounts over some averaged 
time period.  I doubt that it really cost $20, and if for some reason 
the operators had twice as many tape mount requests, that they would 
charge a lower amount.

Anyway... interesting numbers.  I'd love to see the real cost for a lot 
of things played out, including the development costs.  Of course, many 
companies, telco's especially, like to recoup their costs again and 
again, once an asset is fully depreciated.  They also don't want to 
lose money on assets they put in place, but never needed to use.  Good 
example, is all the physical wires and fiber in this country, and yet, 
we are worse than some third world countries as far as our cell phone 
technology.  Go figure :-)

-Carl


On Mar 8, 2006, at 10:08 PM, John C Green Jr wrote:

> =======================================
>
> Posts to inforoots at computerhistory.org is information known to or the 
> opinions of the poster.  All posts to inforoots at computerhistory.org 
> are archived.  By posting to this list you grant a license for use of 
> this material to the Computer History Museum located in Mountain View, 
> California, USA.
> =======================================
> 1 of 2 is Google.  2 of 2 is telephones.
>
> This one is somewhat off topic as we're a computer list, not
> a telephone list.
>
> Back in 1976, after the 1968 Carterphone case but before the
> breakup of AT&T, I spent 5 months at Bell Canada's main
> data center in Toronto.  I learned something that few knew
> at the time but which makes sense once it's explained.
>
> The cost of telephone service (cost, NOT price) includes two
> main components:
>
> * FIXED COST of having the network infrastructure in place
>   so that you can pick up a handset, get dial tone, dial a
>   number, and have a conversation
> * INCREMENTAL COST of actually making such a call
>
> Basic service back then was around $10 per month plus about
> $0.06 per minute, non peak, for actually making a call.
>
> Ever wonder what the actual COST, not price, of making a long
> distance call was back then.  An incremental call that did not
> break the camel's back and require expanding the network
> infrastructure?
>
> It was under $0.03, regardless of distance, and regardless of
> duration.  And 90% of that was capturing the originating
> number, the destination number, the start time and stop time,
> and putting them on a piece of paper at the end of the month.
> The cost of the call was under $0.003 if accounting wasn't
> necessary.
>
> Given that I wondered when price would be cost based, with
> perhaps $30-$50 per month access charge and unmetered usage
> charge.
>
> It took nearly 40 years of competition before some cell phone
> calling plans started having price track cost; there are also a
> few land line calling plans where price tracks cost.
>
> Regards,
> John Green
>
> _______________________________________________
> inforoots mailing list
> inforoots at computerhistory.org
> http://mail.computerhistory.org/mailman/listinfo/inforoots