Index Home About
From: Floyd Davidson <floyd@ptialaska.net>
Newsgroups: ba.internet,comp.dcom.wan,comp.dcom.xdsl,comp.dcom.telecom.tech
Subject: Re: Alternatives to connecting offices by T-1
Date: 12 Feb 2001 17:33:46 -0900

Lizard Blizzard <nospam@rsccd.org> wrote:
>CHANGE USERNAME to just westes wrote:
>
>> Are there *cheap* alternatives to connecting two offices with a
>> high-speed data line (at least 1.5Mbps) other than T-1?   One

...

>Pac Bell, at lease in my area, runs the local lines into the
>'B' box.  Then it usually goes thru a heavy multi hundred pair
>cable clear back to the central office.  They do not allow a
>crossover in the box for the reasons below.
>
>> If two offices are relatively close to each other, and go
>> through the same aggregation point, it seems like a complete
>> waste of resources to force any data line that connects those
>> two offices to travel all the way to the central office for
>> the area.  For

...

>The telcos are wise to these things.  They were installing 'dry
>pairs' for alarm circuits and found that the users were using
>them for DSL instead.  So they make sure that the line has to
>run all the way back to the central office and then back out to
>the same box to your other location.  They want you to buy DSL
>from them, not from a company that sells you a set of DIY boxes
>for DSL.

I've ranted and raved about "Telephone Company Management"
syndrome for the past many years here in comp.dcom.telecom.tech;
however, while the above does fit the typical mindset of a
typical pointy-haired manager in a telephone company, the fact
is none of what you suggest is true at all.

The move towards eliminating direct point-to-point dry pair
circuits of any kind was begun *years* before xDSL was even
thought of, and for some very sound economic reasons.

The telephone industry has always (or maybe I should say that
this was true when it was THE Bell System, all under one
company) studied ways to shave another 1/10 of one cent off of a
few million items to make some bucks.  Two tone multifreq
signaling is the classic example (especially when they turned
around and charged each customer to use touch tone!).

The advent of digital cross connect systems and the rising cost
of "windshield wiper time", not to mention just the cost of a
craftsman, are the reasons that all circuits necessarily must
pass through a central location where testing is possible
without sending out a truck and an installer.

Whether you buy DSL from the telco or not, they do NOT want
_you_ calling in a trouble report that automatically puts
someone behind the wheel of a vehicle for any length of
unproductive time.  If they could spread that cost out amoungst
all customers, or even just over all the LD calls, then it would
be fine.  But they cannot do that anymore, and if they charge a
rate which permits it, nobody will be able to afford the
service.  Hence they just dropped that service as too expensive
for them to provide.

--
Floyd L. Davidson         <http://www.ptialaska.net/~floyd>
Ukpeagvik (Barrow, Alaska)                 floyd@barrow.com


From: Floyd Davidson <floyd@ptialaska.net>
Newsgroups: ba.internet,comp.dcom.wan,comp.dcom.xdsl,comp.dcom.telecom.tech
Subject: Re: Alternatives to connecting offices by T-1
Date: 13 Feb 2001 09:23:31 -0900

"CHANGE username to just westes" <DELETE_westes@uscsw.com> wrote:
>"Floyd Davidson" <floyd@ptialaska.net> wrote in message
>news:8766ifia5x.fld@barrow.com...

>> Whether you buy DSL from the telco or not, they do NOT want
>> _you_ calling in a trouble report that automatically puts
>> someone behind the wheel of a vehicle for any length of
>> unproductive time.  If they could spread that cost out
>> amoungst all customers, or even just over all the LD calls,
>> then it would be fine.  But they cannot do that anymore, and
>> if they charge a rate which permits it, nobody will be able
>> to afford the service.  Hence they just dropped that service
>> as too expensive for them to provide.
>
>Just a quick comment.  When I have problems with quality of
>service on a voice line, I always have to call the problem in.
>I have never once had such a problem resolve on its own unless
>it was a transitory weather-related problem.

That merely suggests you have less exposure to telephone plant
maintenance than you think.  Lots of problems (maybe even most)
are solved without the customer even being aware that a problem
existed.

>   When I do call in such problems, PacBell always rolls a truck.

That is certainly not true.  Problems are screened and whatever
testing can be done from the CO is done first.  Then a truck may
well roll, but in a well defined direction, with some explicit
knowledge of what the problem is and what an expected repair
action will likely be.

>I have never once seen a quality of service problem on a voice
>line resolve itself from the central office.

I have seen a few bazzillion examples.  It is *common*.

>  I have definitely had PacBell tell me that a problem was due
>to internal equipment, but when the problem is theirs I have
>never seen it resolve from the central office.

So you are suggesting that the equipment they install in their
CO's is 100% fault free over its entire lifespan.  Logically,
that is not likely to happen.

>Now, a voice line is much less sensitive to quality issues than
>a data line.

That is not true.  In fact, the opposite is true.  Noise
immunity above a given threshold is the primary advantage of
digital systems.

>  Given my experience with voice, do you really
>expect me to believe that data quality issues on a line that is
>maintained at a PacBell central office are going to get A)
>identified at all, and B) resolved on my behalf, in a way that
>is transparent to me as a customer?

Given what you have described as your experience, I can only
determine that you are not able to determine if that is true or
not.  You have no way to know when PacBell identifies a problem
and resolves it without involving the customer.  Hence you will
observe zero instances of that regardless of what the true
numbers are.

>It is very clear to me that no one is in a better position than
>the customer to determine that there is a quality of service
>issue, IF the customer is empowered with *information*.

Generally the customer pays the provider specifically to avoid
having to interpret and monitor such information.  They do that
because the ability required is expensive and for them it is
cost effective to buy it from the provider.

>  In the
>case of a voice service, you cannot hide the information.
>Believe me, I'm sure that PacBell wishes it could cut off our
>ears or at least blast some dynamite near us whenever it
>installs a voice line, but luckily it hesitates to do so due to
>advice from its lawyers. :)

If  you really believe that your ears tell you more about what
is going on with a voice circuit than PacBell's test equipment...
well, you are mistaken.

>To be cynical, the reason for maintaining quality of service
>data for data lines in a central office is to hide that
>information from the end user.

Why would any telco increase the cost of service due to added
overhead for service that 99 percent of all customers do not want?

>You cannot complain about what you do not understand and do not
>perceive and measure.  Who cares if your router is losing 5% or
>more of all packets due to line quality.  PacBell doesn't take
>responsibility for IP, and they don't let you see what you need
>to see to understand why you have the degradation.  So PacBell
>constrains the problem in a way that leaves it very
>self-satisfied that it is taking care of the customer, but
>which leaves the customer feeling that PacBell doesn't take
>care of providing a solution to the customer's problem.

I can't speak to that specifically because I know little about
PacBell itself.  I would not be surprised to find that what you
say is exactly true!

>So, here is an original thought: let's put the quality of
>service data out in the open, for both the customer and PacBell
>to see equally.

You are free to purchase test equipment from T-Berd, Sunrise, or
whoever...  :-)

>   Yes, I'm
>sure it's PacBell's greatest nightmare, but it's the only arrangement that
>will ever result in the customer getting a high quality data circuit.
>PacBell may have to start charging for services differently, but as long as
>I can get the service from other CLECs, that's fair to all concerned, and
>creates a competitive environment in which all of these companies start
>trying to deliver shrink wrapped, and useful, data services.

I certainly do not want to defend PacBell specifically in any way.  If
you have other options, *use them*.

--
Floyd L. Davidson         <http://www.ptialaska.net/~floyd>
Ukpeagvik (Barrow, Alaska)                 floyd@barrow.com

Index Home About