From: floyd@hayes.ims.alaska.edu (Floyd Davidson)
Subject: Re: The March of Progress
Organization: University of Alaska Institute of Marine Science
Date: Sat, 23 Nov 1991 11:03:36 GMT

John Higdon <john@zygot.ati.com> writes:

> And in the next article, dill@ux1.cso.uiuc.edu (Dill) writes:

>>     Is it not the case that AT&T would not bill for a call until after
>> the first second or two or five?  I remember hearing that somewhere.

> This was a technical limitation in the old mechanical (including
> crossbar) switches. It took up to several seconds for such equipment
> to recognize supervision from the far end and as such would take that
> long to start the billing record. This was not done as a courtesy or
> to provide a "grace period". It was merely a technical limitation.

It may or may not be true, but I'm lead to believe that at least well
into the '60 it took more like 30 seconds to start billing.  And that
it essentially was a "grace" period.  That went by the way when people
figured out how to record something and play it back faster, thus
being able to transfer a significant amount of information in less
that 30 seconds.  I'm sure that data communications via modems put the
final touch on any significant "grace" period.

However, apparently in some places (perhaps all?) regulations require
at least 2.0 seconds of off hook supervision before billing begins.  I
don't keep up with regulations, but the DMS-200 documentation states
that the parameter is set by default to be 2.08 seconds to comply with
any such 2.0 second regulation.  (The parameter may be set from .16
seconds to 40.8 seconds in .01 second increments on a DMS switch.)

>>     Recently, I noticed that I have had a large increase in one minute
>> calls.  So, last month I tried making a few calls and hanging up the
>> moment the other side picked up (say it was an answering machine or
>> something to that effect).  Sure enough, these one minute calls showed
>> up on my bill.  Is AT&T trying to dig out a few pennies or am I just
>> wrong?

> Modern electronic equipment can recognize supervision almost
> instantly.  Added to this is the conversion from inband signaling.
> Your observations are correct, but your assumptions regarding cause
> and purpose are not warranted. Yes, the grace periods are over. But
> remember, AT&T has no control over what the originating switch is.
> That equipment belongs to the LEC and when call timing begins depends
> on the local switch. Even when it does not, as in the case of some
> OCCs, billing will generally begin immediately because most equipment
> is now capable of it.

Call timing for billing purposes is a function of whichever switch
does the AMA recording, which is usually the toll switch (ie. AT&T).
That switch inserts whatever delay (which defaults to 2.08 seconds on
DMS-200 toll switches) there is between called party off hook
supervision and the recorded call start time.  (I don't really think
that the length of time it takes to recognize supervision has much to
do with it.  The two second delay is to avoid false supervision.)

I have no idea if AT&T (or Alascom) or any OCC is required to provide
a 2.0 second delay.  I also have no idea how that delay is handled in
AT&T ESS switches.  I do know that Alascom does in fact use the
default timing of 2.08 seconds.


Floyd L. Davidson   | Alascom, Inc. pays me, |UA Fairbanks Institute of Marine
floyd@ims.alaska.edu| but not for opinions.  |Science suffers me as a guest.