TELECOM Digest     Wed, 4 Jan 95 08:33:00 CST    Volume 15 : Issue 5

Inside This Issue:                        Editor: Patrick A. Townson

    Re: Protest of New Compuserve-Unisys GIF Usage Tax (Jack Hamilton)
    Re: What is a T1 Line? (Joseph H Allen)
    Re: Britain-Japan Fiber Cable (Wally Ritchie)
    Re: Finland Data Transmission (Wally Ritchie)
    NANP 800 Numbers From the UK (Clive D.W. Feather)
    Cellular Billing Services (Raymond S. VanderBok)
    Cell One NY/NJ Eliminates Daily Roam Charges (Stan Schwartz)
    Summary: Telecom Texts (David P. Wiltzius)

TELECOM Digest is an electronic journal devoted mostly but not
exclusively to telecommunications topics. It is circulated anywhere
there is email, in addition to various telecom forums on a variety of
public service systems and networks including Compuserve and America
On Line. It is also gatewayed to Usenet where it appears as the 
moderated
newsgroup 'comp.dcom.telecom'. 

Subscriptions are available to qualified organizations and individual
readers. Write and tell us how you qualify:

                 * telecom-request@eecs.nwu.edu *

The Digest is edited, published and compilation-copyrighted by Patrick
Townson of Skokie, Illinois USA. You can reach us by postal mail, fax 
or phone at:
                    9457-D Niles Center Road
                     Skokie, IL USA   60076
                       Phone: 708-329-0571
                        Fax: 708-329-0572
  ** Article submission address only: telecom@eecs.nwu.edu **

Our archives are located at lcs.mit.edu and are available by using
anonymous ftp. The archives can also be accessed using our email
information service. For a copy of a helpful file explaining how to
use the information service, just ask.

**********************************************************************
***
*   TELECOM Digest is partially funded by a grant from the              
*
* International Telecommunication Union (ITU) in Geneva, Switzerland    
* 
* under the aegis of its Telecom Information Exchange Services (TIES)   
* 
* project.  Views expressed herein should not be construed as 
represent-*
* ing views of the ITU.                                                 
*
**********************************************************************
***

Additionally, the Digest is funded by gifts from generous readers such
as yourself who provide funding in amounts deemed appropriate. Your 
help 
is important and appreciated. A suggested donation of twenty dollars 
per
year per reader is considered appropriate. See our address above.

All opinions expressed herein are deemed to be those of the author. 
Any
organizations listed are for identification purposes only and messages
should not be considered any official expression by the organization.
----------------------------------------------------------------------

From: jfh@crl.com (Jack Hamilton)
Subject: Re: Protest of New Compuserve-Unisys GIF Usage Tax
Date: 3 Jan 1995 23:28:03 -0800
Organization: CRL Dialup Internet Access  (415) 705-6060  [Login: 
guest]


There have been many articles on this topic posted in the 
comp.infosystems.
www.* groups,

You should read all these articles yourself, but a reasonable summary, 
I
think, is this: 

- Compuserve was charged by Unisys with a patent violation for the LZW 
  compression algorithm used in GIFs.  Compuserve had used the 
algorithm 
  believing it was in the public domain, but in fact Unisys had 
applied 
  for a patent.

- Compuserve decided to license the technology rather than fight the 
patent.  

- Compuserve doesn't seem much happier about this than anyone else 
(except
  the people at Unisys, I guess).  

- Unisys clearly seems to be the villain here.

- Compuserve has to pay royalties to Unisys.

- The patent doesn't apply to GIF files themselves.

Here's part of a message from Tim Oren, who claims to be a vice-
President
of Compuserve. 

  CompuServe has not and is not asserting a proprietary interest in 
the GIF
  spec.  Even if we wanted to, there are enough sharp attorneys here 
who could
  remind us that it has long been publicly disclosed with no patent 
filing
  or attempt to assert other rights.  However, in developng GIF, we 
did
  use the LZW compression scheme.  Unknown to us at the time, Unisys 
had
  filed for a patent on that algorithm, even though we found it in a 
public
  source.  We were approached regarding the matter long after GIF had 
been
  widely released on CompuServe and to the market at large.  
Regardless of
  what you may think about software patents in general, or the tactic 
of
  waiting until substantial infringment has occured, we were in an 
infringing
  position and had no option to seek a license.  Since we have a 
substantial
  base of developers who were not only creating clients to Compuserve,
  but had used GIF in good faith in their own products, we also needed 
to
  in some way protect their interests by developng a pass through 
license.

  The text of that license, or portions of it, are what's circulating
  around the nets.  Because it has been taken out of context, many are 
taking
  it as an assertion, by CompuServe, of an intent to prosecute 
proprietary
  rights in GIF against all users, including developers of Web clients
  and other software.  This is not the case.  CompuServe has no intent
  of pursuing rights in GIF in such a fashion, and I am writing this 
with
  the knowledge and consent of our CEO.  Unisys, of course, may follow
  whatever they see as their best interests in the matter.  For better 
or
  worse, as a patent attorney can tell you, selective enforcement is 
allowed
  by the letter of the law.

  Tim Oren
  Vice President
  CompuServe

and more from Mr Oren:

  Re GIF, I can talk about it, and if the impression being left is 
that
  we are trying to make big bucks, I definitely need to talk about it,
  because we've left the wrong impression.  Here's the story, which 
you
  CAN repeat, WITH my name attached:

  GIF was originally developed at CompuServe by Steve Wilhite, who
  currently works for me.  As part of it, he used with the LZW
  compression scheme, which had been openly published by Unisys 
engineer
  in a journal.  A number of other developers picked up and used LZW 
as
  well.  None of us knew that Unisys had filed for, and eventually
  received, a patent on the LZW scheme.  I believe this is called a
  'submarine' patent - it can surface and get you later.  We were got.

  Unisys proposed an infringement suit, and we had no recourse but to
  settle.  We are paying licensing fees in a manner which IS a
  nondisclosure item.  One of the things we needed to be able to do is 
to
  'pass through' a license embodying both LZW & GIF to those 
developers
  who create their own client programs to CompuServe, such as TAPCIS 
and
  Mac Nav, since they 'practice' LZW as well.  The reason that GIF is
  included as a conditional in such licenses is that we can't pass
  through an unrestricted LZW license, and the reason there is money
  involved is that we in turn have to pay Unisys.

  If anyone is taking the impression that we are asserting proprietary
  rights in GIF additional to the LZW patent, that is wrong.  Neither 
are
  we attempting to assist Unisys in enforcing their patent with 
respect
  to non-CompuServe environments, such as the Internet, though 'buying
  into' our license would be one way of Internet based vendors in
  avoiding possible action from Unisys.

  This is far more headache than it is worth, believe me, and we are
  actively engaged in looking at migration strategies that will get us
  and our customers off the hook.  Our reputation has been damaged by
  being an unwitting partner to Unisys during those 7 years of
  encouraging proliferation, and we are not happy about it.  (I'm sure
  Stallman - rms - could find a moral in here somewhere... )

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

Jack Hamilton  jfh@crl.com  KD6TTL  '92 K75RTA  co-moderator, 
sci.med.aids

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

From: jhallen@world.std.com (Joseph H Allen)
Subject: Re: What is a T1 Line?
Organization: The World Public Access UNIX, Brookline, MA
Date: Wed, 4 Jan 1995 08:01:22 GMT


In article <telecom15.4.4@eecs.nwu.edu>, James Carlson 
<carlson@xylogics.
com> wrote:

> I'm afraid you'll have to re-read what I wrote.  There is no such
> thing as "cross-talk" at the *digital* level.  One DS0 channel 
cannot
> affect another within a T1 line; there just is no path for this to
> happen.  Any signal at all can be carried on a DS0, and the others
> will be blissfully unaware that it exists.  They are separated in
> time.

For a DS0 as defined as a TDM channel on a T1, this is true (excluding
inter-symbol interference on a bad line).  The original poster was
concerned about a filter on a 56K copper pair, which prevented it from
being used at any higher rate.  In this case there is filter to
prevent cross-talk between 56K copper pairs.  An entire T1 can also go
on a copper pair, but not with that filter.  The fact that the line is
a DS0 which eventually gets multiplexed onto a T1 also limits his
bandwidth. If this is not the filter which the poster was referring
to, then pardon me.

>>> In fact Pulse Code Modulation (digital) without any error 
detection
>>> has almost the same interference properties as FM radio.  If two 
FM
>>> stations are close, there will be noise.  If they are really 
close,
>>> you'll hear both stations at once and lots of noise.

> The two are very different.  PCM is a digital signal; if you can
> recover it from the noise on the line, you'll get a "perfect" copy 
of
> the data.  Line hits will produce a decidedly non-analog type of
> noise, with random data points being generated.  To produce
> traditional "cross-talk" effects, you'd have to do an arithmetic
> average on the values of the data points between two lines.  This is
> highly unlikely to occur.

> If you had two AMI lines next to each other, you'd need to have the
> extraneous cross-talk signal exceed the pulse detection threshold of
> the receivers in order to be able to detect *anything* at all.
> Anything below this value would simply be invisible at the digital
> level.

If there are two PCM signals going at the same rate with no error
detection, no sequence randomization, and out-of-band syncronization,
and there is enough interference to mess with the detection threshold,
you will here actual cross-talk.  It will be very bad, and the
original signal will also be almost completely munged.  Yes, this is
different from linear additive cross-talk, but it's still cross-talk.

> FM, due to the nature of analog demodulation techniques, exhibits a
> capture effect.  If two signals are broadcast on the same frequency,
> the "louder" signal will be heard to the exclusion of the "quieter"
> signal.  You won't hear both stations at once.  If the received
> amplitude of the two signals are varying with respect to each other,
> then all sorts of interesting effects (depending on the type of
> demodulation used) occur, but hearing both at once isn't one of the
> options.

This "capture" effect has nearly the same effect as two PCM signals on
top of each other.  The stronger one will get through if the weeker
one doesn't mess with it too much.  If they have the same signal
levels, you'll have real cross-talk plus lots of noise (to the point
where they are both nearly indistinguishable, same as PCM).  This
really happens, try it.

>>> There's also a ~100KHz low-pass filter between the A/D converter 
and
>>> the line to limit the edge rate of the digital pulses.  Cross-talk 
is
>>> more severe at higher frequencies, so limiting the unnecessary 
high-
>>> frequencies caused by fast edges helps reduce cross-talk. 

> Any filter placed in the DS0 data path after quantization and before
> multiplexing has no effect on the bandwidth of the signal that the
> user of that DS0 sees.  This has nothing to do with the poster's
> original statements.

It does, because the original poster was concerned about that
bandwidth of the actual copper pair that came to his house.  It's my
understanding that an ISDN DS0 is 56K digital line on a copper pair
between your house and the local office, not a shared T1 (although a
multiplexer at the local office my put it on a T1).


jhallen@world.std.com (192.74.137.5)  Joseph H. Allen

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

From: writchie@gate.net
Subject: Re: Britain-Japan Fiber Cable
Date: 4 Jan 1995 06:31:23 GMT
Reply-To: writchie@gate.net


In <telecom14.474.16@eecs.nwu.edu>, nagle@netcom.com (John Nagle) 
writes:

> wrf@ecse.rpi.edu (Wm. Randolph U Franklin) writes:

>> AT&T will build a cable from Britain to Japan for $1.2G.  It'll be
>> 17,000 miles long, 5Gbps, and carry 320,000 "voice and other 
messages".  
>> That looks like only 16Kbps per circuit (which looks low).  The
>> current longest cable is a 9,000 mile one from France to Singapore,
>> completed a year ago.

>> Fun math: That works out to a capital cost per circuit of only 
$3750.
>> Assume that a phone call from Britain to Japan costs $2/minute.  If
>> all 320,000 channels were in continuous use, then the cable would 
be
>> paid for in the first 31 hours.

>> Alternatively, if we assumed that the cable is good for ten years,
>> or 100,000 hours, then amortizing the capital cost would be three
>> cents per hour, or $5e-4/minute.  This is a factor of 4,000 less
>> than the price of the call.

>      The numbers for the newer transantlatic cables look like that,
> too. You really should be able to buy a full-time transatlantic
> circuit for about $100/month, and at the rate cable is being laid, 
you
> probably soon will.

>      Not having to acquire property rights is a big win. 
Fortunately, 
> the UN didn't think of this for the Law of the Sea conference.

There is no question that the "cost" to the carriers of the
international circuits is declining toward zero. That cost, however,
has very little to do with what you pay for an int't private-line or
switched service. An international circuit is two "half circuits" with
the price at each end completely controlled by the carriers at each
end, one or both of which is normally a monopoly PTT. Only US/Canada
and US/UK circuits are priced at anything even reasonably related to
cost. Until PTT's discover what price elasticity means, or they are
forced by large private bypass users, tariffed rates are unlikely to
reflect the low costs actually involved.

The U.S. Model is informative. For interstate long distance the
"costs" are $0.03 originating access (paid to LEC), $0.03 terminating
access (paid to LEC), and $0.01 or less TOTAL operational cost for the
IXC.  Because the IXC has to pay the LEC's the $0.06, it has about
$0.07 total cost resulting in a retail pricing on the order of $0.11 -
$0.16. The IXC of course has to absorb credit risk on all of the true
cost (including the LEC which gets its money in any case. The IXC also
has marketing and other costs. Nevertheless the true cost of long
distance is on the order of a penny a minute and that includes the
switches. If access costs were reduced to the same magnitude as the
long distance, the total retail price would be on the order of a
nickel a minute. Intrastate is even worse. I actually pay less to call
Ft. Lauderdale to London than I pay to call from Ft. Lauderdale to
Miami.

The fact is that most PTT's subsidize their large and politically
powerful employee bases by maintaining the highest possible overall
charges for international calling both in collection rates (charged to
those in their country) and in accounting rates (charges to U.S. or
other foreign carriers). The same true for private lines. In fact the
"price" guideline for a private line is oftern based on 9000 minutes
of switched usage, no relataion to cost.


Wally Ritchie    Ft. Lauderdale

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

From: writchie@gate.net
Subject: Re: Finland Data Transmission
Date: 4 Jan 1995 06:50:24 GMT
Reply-To: writchie@gate.net


In <telecom15.1.10@eecs.nwu.edu>, jackp@telecomm.admin.ogi.edu (Jack
Pestaner) writes:

> We have been communicating to a site in Finland with autoranging 
14.4k
> modems.  On a good day we can run at 9600, but typically at 2400.  
We
> have tried AT&T, MCI, and IDB (all are direct digital connections
> through our PBX), but all seem to be extremely variable.  We use
> NetBlazer modems, same model, on each end.

> This is really expensive, and we want to move to a more reliable
> service, as we expect to have longer hold times of three to five 
hours
> a day.  I checked on a 56k DDS, but cost was about $9K per month.

> Are there any satellite solutions, or packet solutions that anybody
> knows of?  BTW, we also tried x.25 from Sprint, but service went 
down
> often, and Sprint just has the WORST customer service for problem
> solving.

You might want to try different modems including some newer V.34's You
might also investigate the nature of the analog loops at both your end
and the Finland end. V32bis and V34 depend on effective cancellation
of echo. Not all modem models are equally effective with circuits that
have long delays. The types of echo cancellers and echo suppressors
used on individual connections can also be a significant factor. Since
you will be spending several $K per month, it should be worthwhile to
investigate different modems and possibly some conditioning of the
analog loops in Finland. How is your voice quality? Are you getting
satellite circuits all the time?. What happens US/UK with the same
modems? What happens Finland/UK.

With the money your spending, the capital costs of the best modems
money can buy are insignificant.


Wally Ritchie    Ft. Lauderdale, Florida

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

Subject: NANP 800 numbers from the UK
Date: Wed, 4 Jan 1995 12:57:52 GMT
From: Clive D.W. Feather <clive@sco.COM>


All of a sudden, I can now dial 1-800 numbers from the UK. There is a
few seconds of silence, and then an American female voice tells me
that this call will *not* be free, but charged at standard direct-dial
rates.  If I don't want to pay, I should hang up now.

Just for fun, I tried 1-800-MY-ANI-IS. It told me: "702 000 5555" !


Clive D.W. Feather     | Santa Cruz Operation    
clive@sco.com          | Croxley Centre          
Phone: +44 1923 813541 | Hatters Lane, Watford   
Fax:   +44 1923 813811 | WD1 8YN, United Kingdom 

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

From: rvb@iti.org (Raymond S. VanderBok)
Subject: Cellular Billing Services
Date: 4 Jan 95 13:02:09 GMT
Organization: Industrial Technology Institute


Are there service companies that provide custom billing for cellular
service?  I would like to be charged for service in a way different
from what my cellular provider offers.


Ray VanderBok
Metalforming Manager
NIST / Midwest Manufacturing Technology Center
(313)769-4131    internet: rvb@iti.org

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

Date: Tue, 3 Jan 1995 22:35:58 EST
From: Stan Schwartz <stans@panix.com>
Subject: Cell One NY/NJ Eliminates Daily Roam Charges


It was buried in small print in the copy of their ad in Monday's {New
York Newsday}, so I called customer service and verified it.  As of
1/2/95, Cell One NY/NJ (00025) has FINALLY followed NYNEX Mobile's
lead and eliminated the $3.00 daily "nag" roamer charge everywhere in
North America.  This applies to both NACN and Non-NACN cities (it was
unclear in some cases whether some NACN cities charged the fee).  At
any rate, the only charges for roaming outside the NACN are 99 cents
per minute airtime (still higher than NYNEX, for the most part).
CellOne's customer service rep explained that some areas are 83 cents
per minute instead of the 99, but they are few and far between - one
should assume that it will be 99 and be pleasantly surprised if the
bill is 16 cents less (per minute).

A small celebration is in order!


Stan

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

From: wiltzius@anduin.ocf.llnl.gov (David P Wiltzius)
Subject: Summary: Telecom Texts
Date: 4 Jan 1995 00:00:06 GMT
Organization: Lawrence Livermore Nat'l Lab.


I had several requests to post a summary so here goes.

I haven't found a good textbook in telecom, but I think Newton's
Telecom Dictionary is a very useful substitute.  Because it is
structured as a dictionary, one must piece together information from
various entries, but I did pretty well on your test sentence. All the
technical terms in your sentence were in Newton and they clearly
pointed me toward the SONET entry, which is almost two pages long, and
did a decent job of explaining your sentence, I think.

I teach a course on information and communication technology (for
industrial engineers); I use a collection of up-to-date articles on
uses of telecom (Network World is one great source) and have the
students use Newton to help them read these articles.

I saw a book in Barnes and Noble called "Digital Telephony", Second
Edition, by John Bellamy.  Publisher is John Wiley, (c) 1991.

IBSN  0-471-62056-4

After perusing the book, I came to the conclusion that it contains
much of the information I am looking for (the same sort of thing you
are looking for).

The first chapter covers analog systems.  The rest is digital.  There
is a chapter on SONET.

I am still debating whether or not to get it (it was $75).  I am
interested but I could also spend $1000 on software engineering books
(my occupation).

Consider it requested.  BTW, I bought some of the Telephony BASIC
series.  Haven't had a chance to look at them yet but they are THIN.
I suspect they are VERY basic and not a good bargain.  Happy Holidays,
(Ed: I bought two of them (SONET and Wireless) months ago and I think
they are worth their modest price.  They are basic, as advertised.)

McGraw-Hill, Delran, NJ 08075, publishes a series of reference books
called Datapro. The series covers everything from how modems work, T-1
history, to vendor and product analysis.  William Stallings has
written several books on communications media and equipment.  Titles
slip my mind but I'm sure you should be able to find something by
author search.

-------------------Start off with "Digital Telephony" 2nd Ed. by John
Bellamy ISBN 0-471-62056-4.  He has a chapter on SONET that will give
you a taste.  Since the field is still emerging, many issues with
SONET and SDH are still evolving.

------------------- RADIO SHACK CARRIES A BOOK CALLED "UNDERSTANDING
TELEPHONE ELECTRONICS" WHICH EXPLAINS SUBJECTS FROM THE SIMPLE POTS
TELEPHONE TO CENTRAL OFFICE EQUIPMENT.  TELLABS, INC. PUTS OUT A TEXT
BOOK COMPLETE WITH TESTS AT THE END OF EACH CHAPTER.  HOPE THIS HELPS.
IF YOU HEAR ABOUT OTHERS PLEASE LET ME KNOW.

------------------- You're welcome! I recently purchased an excellent
telecom book from Artech House Inc.: "Service Management in Computing
and Telecommunications." Artech specializes in telecom material. Their
number is: 617-769-9750, call for a catalog. I've found them pleasant
to do business with, but their prices, as with Computer Literacy, are
full list (read: high).

Speaking of price, if you're in the San Jose area, there is a large
store near the corner of Hilsdale and Camden, across Hillsdale from
Target, that buys close-out books from publishers. They sell most for
under $5.00, and many for $3.00! Many books are college texts, and on
some very good subjects. I was there this afternoon and picked up the
"Artifial Inteligence Handbook ($3.00), Spencer Johnson's "Yes or No"
($3.00), "Voodoo Dos" ($8.00), Jimmy Carter's "An Outdoor Journal"
($3.00), Stanley Davis' "Managing Corporate Culture" ($2.00), and the
"Pocket Guide to Phrasal Verbs" ($2.00). These are all hardbound
editions. I didn't see any telecom material, but I have in the past.
the place is huge, and I didn't look at everything. It's definately
worth a visit if you're in the neighborhood, and you have the time.

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

End of TELECOM Digest V15 #5
****************************

