A Beginner's Guide to the Internet/INTERNET.ISO/text/telediag/tc13-084.txt

Text File | 1996-05-06 | 18.3 KB | 469 lines

TELECOM Digest     Fri, 12 Feb 93 01:02:00 CST    Volume 13 : Issue 84

Index To This Issue:                     Moderator: Patrick A. Townson

    Re: German PCN License (E1) Given to VEBA and Thyssen (Lutz Albers)
    Re: Prodigy <> Internet Gateway Almost Finished (Kauto Huopio)
    Re: 1ESS and CNID (Arnette Schultz)
    Re: PacBell IntraLATA Rate Ripoffs (Glenn McComb)
    Re: Meet Me at the Power Line (John Nagle)
    Re: Using Fax Machine + Fax Modem as a Scanner (Jarom Hagen)
    Re: SS7 / CID Question (Mark Baker)
    Re: AT&T Are You Listening? (John Higdon)
    Re: Human Factors For Speech Recognition Systems (Charles Hoequist)
    Re: What The Heck is 4,941,976.667 Hz Good For? (Bob Clements)
    Followup: Rochester Tel Wants to Split (Phillip Dampier)
----------------------------------------------------------------------

Date: 11 Feb 93 15:44:58 EST
From: Lutz Albers <[e-mail]>
Subject: Re: German PCN License (E1) Given to VEBA and Thyssen


In <TELECOM Digest V13/076>  [e-mail] (H. Shrikumar) wrote:

> I wonder what the competitive environment is?

> Would E-plus and E-star compete? What about D1 and D2? If not
> today, are there plans? Or is there a zonal split?

  [stuff deleted]

> PS: Wow! The services and companies providing them are also named
> so methodically!! An admittedly irrational conclusion re. the
> possible "non"-competitiveness springs in my mind ! :-) Or is it just
> the fabled German perfection at work ??

Well, I think an explanation of the German celluar networks is in
place here.

The oldest working celluar network is the B/B2 network, which is
operated by the DBP Telekom (DBPT) (our dear beloved telecom service
provider ;-). This is an analogue network with no roaming (you need to
know the zone where the called party is located). This network is to
be dismantled around the end of 1994.

The successor of the B network is the C network, which is also
operated by the DBPT. Again this is an analogue network. For this
network you don't need to know the exact location of the called party,
just calling 0161 + number will get you through. According to
information of the DBPT this network uses cells with a radius of
approximatly 20 km and can serve up to 800.000 customers.

A few years ago the German Ministry for Telecommunications decided to
give licenses for a new digital celluar network (based on the new
European GSM standard) to the DBPT (D1) as well as to the private
consostium Mannesmann Mobilfunk (D2). Both networks went into service
last year (at least in the metropolitan areas and along the freeways).
To understand the additional introduction of the E network one needs
to know that both provide plans to cover nearly all areas of Germany.
BTW, this network will provide international roaming in countries with
GSM networks.

Last year the German Ministry decided to give an additional license for
an additional digital network which will base also on the GSM
standard, but will use smaller units with less power. The DBPT was
barred from building such a network so two consortiums names E-plus
(Veba, Thyssen et al) and E-star (BMW et al) were bidding for the
license which was given finally to E-plus. This network will not cover
all of Germany.

So there will be now four celluar networks competing (two of them run
by the DBPT).


Lutz Albers          InterNet:   [e-mail]
Reginfriedstr. 10    CompuServe: 100022.1102
8000 - Muenchen 90   Phone:      [phone]
Germany              Fax:        [phone]

------------------------------

From: [e-mail] (Kauto Huopio)
Subject: Re: Prodigy <> Internet Gateway Almost Finished
Date: Fri, 12 Feb 1993 05:31:32 GMT
Organization: Lappeenranta University of Technology, Finland


In article <[e-mail]> I wrote:

> So, it might be even possible to get a telnet connection to BIX?

I got an answer from Sarat Vemuri ([e-mail]) that
x25.bix.com is the way to telnet to BIX. The DNS (Domain Name Service)
gives these addresses to x25.bix.com:

x25.bix.com     internet address = 192.80.63.1
x25.bix.com     internet address = 192.80.63.2
x25.bix.com     internet address = 192.80.63.3
x25.bix.com     internet address = 192.80.63.4
x25.bix.com     internet address = 192.80.63.6

So, there seems to be plenty of capacity. Someone with a BIX account,
give x25.bix.com a try and report to comp.dcom.telecom the results!!

Sarat notes the charge: "Charges are $1/hr billed to your BIX
account."


Kauto Huopio ([e-mail])
Mail: Kauto Huopio, Laserkatu 3 CD 363,  SF-53850 Lappeenranta,Finland

------------------------------

From: [e-mail]
Date: Thu, 11 Feb 93 18:01 CST
Subject: Re: 1ESS and CNID


In Telecom-Digest: Volume 13, Issue 76, Message 11 of 16 DREUBEN@
EAGLE.WESLEYAN.EDU (Douglas Scott Reuben) writes:

> A number of question have come up regarding SS7 and CID:
> 1. Can a 1ESS switch display CID? Are there any software upgrades/hardware
> upgrades which allow it to do so?

Nope.  1ESS does not support SS7 or LASS features in any form.  This
includes Caller Identification features such as Calling Number
Delivery or Calling Name Delivery.  The poor old 1ESS just doesn't
have the memory space to support the extensive software for these
advanced features.

The 1ESS does not support SS7.  Never has, never will.  The 1A ESS is
a different story.  It does fully support LASS and SS7.

Since a 1ESS can not support SS7, Calling Party Number (a.k.a. caller
identification) is never sent from a 1ESS.  A 1ESS will of course
signal ANI (billing number) information using the (older) Equal Access
Multi-frequency signaling protocol. (See FAQ to understand the
differences between CPN and ANI.)

> Every 1ESS that I've tried to get CID info from fails (ie, you call
> from the 1ESS to a 5ESS that has CID).  Is the Telco doing this
> intentionally, ie, so that customers on the 1ESS who can't hit *67 to
> make their calls private will not have to show their number to
> EVERYONE they call, or is it more of a technical reason why one can't
> get CID from a 1ESS?

I assume you receive an "Out of Area" or "Number unavailable"
indication.  Technical reason.  1ESS does not support SS7 or LASS
(including the "privacy" features).  It doesn't need the privacy
features, because an originating party's number (CPN) is never
transmitted from a 1ESS.

Another reader from Canada remarked on what they assumed was Pacific
Bell blocking Caller Identification because their display for calls
originating from California was "Private" not "Out of Area" (or number
unavailable, etc).  It is possible. Several switches support an "all
numbers private" indication.  I'm not sure if this is a Bellcore
requirement, or simply a fact of Calling Number Delivery evolution.
(Some states that don't permit calling number/name delivery wanting to
insure that numbers are "never" delivered, even out of state.)


Arnette Schultz [e-mail]
(Opinions and inaccuracies are all mine!)

------------------------------

From: [e-mail] (Glenn McComb)
Subject: Re: PacBell IntraLATA Rate Ripoffs
Organization: Netcom - Online Communication Services  (408 241-9760 guest)
Date: Fri, 12 Feb 1993 06:38:58 GMT


Oops. Sorry to hear that your credit card call got hit for an
additional 0.75

When I originally posted this a while back, I wanted to be able to
access MCI from my primary number, not just credit card access.  I
mean: 10222+ dialling.

BTW, Execuline, which I now use for 415/510 calls from 408 charges
0.12/min no matter what time of day.  Thank you speed dial!


Glenn A. McComb        [phone] ofc * 725-0222 fax
McComb Research        PO Box 220 * Cupertino, CA 95015
gmccomb @ netcom.com                MHS: glenn @ mccomb

------------------------------


From: [e-mail] (John Nagle)
Subject: Re: Meet Me at the Power Line
Organization: Netcom - Online Communication Services (408 241-9760 guest)
Date: Fri, 12 Feb 1993 02:02:42 GMT


[e-mail] writes:

In <[e-mail]>, [e-mail]
writes:

> The inverse-square law holds only for point sources.  For a line, the
> corresponding law is straight reciprocal: inversely proportional to
> the distance from the line.

       However, for a multiconductor power line, there is a
cancellation effect from the return current, so the geometry is more
complicated.  The ratio of the distance between the conductors and the
distance to the conductors is key.  When you are far away from a pair
of wires near each other (like an ordinary power cord) the effects of
the two conductors cancel out.  Twisted pairs cancel even better.  So
an analysis based on a single-line model isn't valid.

       Three-phase lines require more analysis, but I think that the
effects of all three lines cancel similarly, since there's no net
electron flow (the current in all three lines instantaneously sums to 0).


John Nagle

------------------------------

From: [e-mail] (Jarom Hagen)
Subject: Re: Using Fax Machine + Fax Modem as a Scanner
Date: 11 Feb 93 18:37:14 GMT
Organization: NPRI, Alexandria VA


[e-mail] (Robert J Barth) writes:

> nHow can I use a standalone fax machine as a scanner?

> I would like to be able to plug my fax machine into my pc fax/modem,
> and press a few buttons, and end up with a fax format file of whatever
> was on the sheet of paper ...

All you need to do is to connect the fax and the modem with phone line
cable, have the fax dial a number to make it look for a fax and then
put your fax modem on manual receive (from the software).

It works for me.


Jarom
*Not paid for and/or endorsed by NPRI.  602 Cameron St, Alexandria VA 22314
(UUCP: ...uunet!uupsi!npri6!jhagen)       (Internet: [e-mail])

------------------------------

From: [e-mail]
Date: Thu, 11 Feb 93 07:59 CST
Subject: Re: SS7 / CID Question


> A number of question have come up regarding SS7 and CID:

> 1. Can a 1ESS switch display CID? Are there any software upgrades/hardware
> upgrades which allow it to do so?

The 1 ESS(tm) Switch simply exhausted its available address spectrum
several years ago to allow new capabilities such as SS7 and CLASS to
be implemented on it.  The 1 ESS processor was designed in the early
60's and has a maximum on 768K words of magnetic card storage for
generic program and office data (both equipment configuration and
line/trunks/routing type translation information.)

The 1A ESS switch does support SS7 and CLASS services.  The 1A ESS is
a 1 ESS with a 1A processor which was designed in the mid 70's and has
a 12Mbyte memory spectrum.

> 2. Assuming a 1ESS can't offer CID to its local subscribers, if it is
> connected with SS7 to a 5ESS or a DMS or in general some CID-capable
> switch, can the 1ESS send out the CID information?

1 ESS can't do SS7 therefore it can not send out CID.

> Every 1ESS that I've tried to get CID info from fails (ie, you call
> from the 1ESS to a 5ESS that has CID).  Is the Telco doing this
> intentionally, ie, so that customers on the 1ESS who can't hit *67 to
> make their calls private will not have to show their number to
> EVERYONE they call, or is it more of a technical reason why one can't
> get CID from a 1ESS?

Since 1 ESS does not support SS7 or CLASS it does not even support the
ability for customers to dial *67.  Almost all calls are carried via
MF signaling in and out of 1 ESS switches.


Mark Baker      AT&T Network Systems

------------------------------

Date: Thu, 11 Feb 93 10:41 PST
From: [e-mail] (John Higdon)
Reply-To: John Higdon <[e-mail]>
Organization: Green Hills and Cows
Subject: Re: AT&T Are You Listening?


[e-mail] writes:

> I am thinking about MCI instead.

> I don't want to switch carriers.

Then don't. Use AT&T for whatever AT&T does well, and use MCI for
whatever you perceive to be the better deal from that company. The
concept of "switching" is a matter of brainwashing from the long
distance carriers. Yes, you can only have one PIC or "dial 1"
provider, but you can have as many accounts with as many companies as
you like.

For instance, keep AT&T as your PIC and then sign up with MCI for the
international plan, keeping your MCI account secondary. Then when you
call China (or whereever), prepend '10222' to the call. Frontline
salesslime will tell you that you have to "switch" to take advantage
of a particular plan, but that is 99.9% hogwash.

> So if there are any AT&T people out there who can pass on these
> comments to the powers that be, please do so. You may be losing
> another customer to the opposition.

I have used various plans offered by Sprint and MCI, but I have never
"left" AT&T. There is no need.


John Higdon  |   P. O. Box 7648   |   [phone]     |       FAX:
[e-mail] | San Jose, CA 95150 | 10288 0 700 FOR-A-MOO | [phone]

------------------------------

Date: Thu, 12 Feb 1993 15:08:00 +0000
From: Charles (C.A.) Hoequist <[e-mail]>
Subject: Re: Human Factors For Speech Recognition Systems


This is just to add a few bits and pieces to Norm Tiedemann's
response on this topic.

While it is true, as he wrote, that large-vocabulary systems (say, >
1000 words) by and large haven't been trained for telephony speech,
this isn't because vendors are perverse.  The data gathering necessary
to do the training even for small-vocabulary systems requires
considerable time and money, and it's not clear what the return would
be. Dragon Systems, for example, focuses on applications like spoken
input to a word processor. That's not something people want to do over
the telephone very often. I suspect these hurdles are even more
daunting for vendors than the problems of telephony speech, nasty as
that is.

Some additional sources of information:

_IEEE Computing_ Magazine for August, 1990 was devoted to "Voice in
Computers", or something similar, and I seem to remember some articles
there on speech recognition in the public network.

Norm Tiedemann mentioned one set of ICASSP proceedings. I would expand
that and say pretty much all of them in the last five or six years
have something on telephony ASR.

More human-factors oriented work is in the ACM CHI (Computer-Human
Interaction) Journal. They also have a yearly convention which
publishes proceedings. In fact, there's a good article in the 1991
proceedings of CHI. That's the name of it, incidentally: 'Proceedings
of CHI'. Sounds like a conference on Eastern philosophy.  The article
is "Should we or shouldn't we use spoken commands in voice interfaces?"

Good luck. Since this is a topic I'm following, I'm interested in
whatever gets turned up.


Charles Hoequist   |Internet: [e-mail]
BNR Inc.           | 919-991-8642
PO Box 13478, Research Triangle Park, NC 27709-3478

------------------------------

Subject: Re: What The Heck is 4,941,976.667 Hz Good For?
Date: Thu, 11 Feb 93 15:20:51 -0500

From: [e-mail]


In TELECOM Digest, "John R. Ackermann" <[e-mail]>
wrote:

[An explanation of what he thought that frequency, 4,941,976.667 Hz,
was all about.]

In private email, we have determined that he was wrong, basing his
response on a mis-remembered but similar number used in some frequency
standards.

So my question still stands:

What would AT&T have wanted that rather oddball frequency for, so
precisely that they bought special customized atomic standards to
generate it?  Repeating: It really is a special for AT&T, not a
consequence of the rubidium standard's internal workings.


Thanks for any wisdom.

Bob Clements, K1BC, [e-mail]

------------------------------

From: [e-mail] (Phillip Dampier)
Reply-To: [e-mail]
Date: Thu, 11 Feb 1993 13:03:46 -0500
Subject: Followup: Rochester Tel Wants to Split


Here are further details on Rochester Tel's proposal, courtesy of a
trade publication for the cable television industry ...


ROCHESTER TEL WANTS TO SPLIT
By Rachel W. Thompson  Multichannel News
February 8, 1993

Anticipating competition, Rochester Tel last week proposed breaking
its New York operations into two pieces.

Under a plan filed with the New York State Public Service Commission
last Wednesday, rate-regulated R-Net would take over the assets and
personnel that support basic switching and transmission functions in
the six-county Rochester area and deregulated R-Com would market both
wholesale and retail services using R-Net's network, which would also
be available to competitors.

Time Warner, Inc. is part owner of FiberNet, Inc., an alternate-access
telephone company that operates in Rochester alongside Time Warner's
Greater Rochester Cablevision.

FiberNet vice president Jim Geiger called Rochester Telephone's
proposal "a structure that the pro-competitive regulators have hinted
at," and "commendable."  But if the company believes it can prevent a
physical overbuild of its facilities through the restructuring, that's
"faulty logic, considering that [the cable company is] about 1,400
miles into [a] fiber overbuild."

GRC, some of whose fiber belongs to FiberNet under a condominium
arrangement, passes 300,000 of the 360,000 homes in Rochester
Telephone's local-access telephone area (LATA) "with a broadband
plant, and 200,000 of them are customers," Geiger said.

"By the time that [Rochester Telephone's plan] could conceivably
become effective," FiberNet will be ready to offer a full range of
telecommunications services, he said.  Asked whether FiberNet would be
ready to start offering basic residential services, he answered, "I
would say that as the rest of the local exchange becomes deregulated,
we're certainly interested in exploring those niches."

FiberNet is a joint venture of Time Warner, Petrocelli Industries and
three individuals, including Geiger, and was founded in April 1990.

Petrocelli Industries is a privately held electric and communications
contracting firm based in New York City that does construction for
cable and telephone companies, including Teleport Communications
Group.  Time Warner is a minority partner in FiberNet, Geiger said.

------------------------------

End of TELECOM Digest V13 #84
*****************************