Goodbye to copper? [Telecom]

Originally various consortiums of other Telcos tendered to provide the FTTN network that the government wanted, but no one came up with a suitable solution so the FTTH network was announced. In that tender process Telstra were excluded basically because of their arrogance in assuming that they were essential in any FTTN rollout, so they put in a tender that blatantly did not meet some basic requirements and then complained when they were left out!.

Just today it has been reported that Telstra may well be involved in part of the FTTH network.

AFAIK the government intends to initially set up the network in partnership with private enterprise, and then gradually sell out of it as time goes on.

Which is the dream for most of us who cannot stand dealing with the dominant telco right now.

Remember 1999, every switch needed a software update, it went with very little trouble.

I disagree with that. It's not a vague statement, it's a statement of actual policy and practice.

If the local telephone companies actually were telling people, "sorry, we don't offer POTS, you MUST buy expensive FIOS instead", there would be a national outcry and widespread nasty publicity.

As others pointed out there, there is a demarc box and they must serve it. That part of the business is still regulated.

Having 'three' carriers is _not_ competition.

Certain services--such as utilities--are such that the capital and infrastructure requrements are so high that it is impractical to have competition.

They tried 'competition' in electricity in some places and got blackouts as a result. They have it my state and the power company significantly cut back its lineman and maintenance staff so as to be cost competitive. So when there's a thunderstorm and lines get knocked down people have to wait much longer for service restoration.

Then why, when I had FIOS installed at my residence last year, the verizon salesweasel specifically told me that once I switched my phone service to the FIOS, I would not be able to switch back. And when the techs came out and installed it, they removed the copper cable that had carried my phone service from the house to the street pole.

--Dale

Did they really? If you're talking about Y2k, most switches were new enough to have that already included. (Y2k was an overrated scare-- many computer systems needed no modification at all, and others needed very little modification. But I knew of one highly vaunted system developed by a "Big 8" accounting firm consultant that was so bad it had to be scrapped altogether. So much for the glory of 'consultants'.

Anyway, in the caller-id/ANI discussion, everyone made a emphatic point of saying how difficult and expensive it would be to modify a switch merely to control caller ID abuse. Changing all the dialing tables, routing codes and associated software, plus everything and anything that handles the phone number, would be a far more massive job.

I really meant the whole conversion process, at all levels, in going from multiple long-standing national currencies to one unified Euro: producing and distributing the physical currencies, taking in the old currency and exchanging for the new, converting vending machines and even trivial things like coin trays, re-pricing retail goods, and all the complexities of adjusting credit card accounts, payrolls, tax calculations, banking transactions, foreign currency exchanges outside the Euro system, and adjusting everything at every level of the entire banking system.

It just seems to me to have been a massive and successful changeover. Maybe it was aided at some levels by the extensive use of post office operated "giro" systems for so many consumer and personal transactions, which I believe is or at least was very common in Europe, and near unknown in the U.S.

Europeans had an advantage: they were used to dealing with multiple currencies, so the Euro could be treated as a "new nation" in their computing systems.

There is a surprising amount of "behind the scenes" technology involved in handling coins and bills. Some of it is in "obvious" places (i.e. ATMs and bill counters in banks) but the bulk of it is "hidden in plain sight". When a currency "conversion" takes place, coin and bill acceptors/ dispensers in vending machines (and pay phones!) have to be changed; if the new coins and bills are physically different dimensions than the old, these changes usually require new/ modified hardware. Cash registers also require changes, including new/ modified tills if bills change size, or a new denomination is added.

In the USA, we have had quite a number of "conversions" in the past decade or three, starting with the "Susie" dollar coin and _two_ conversions of most folding money denominations ("security" upgrade then Monopoly-money coloring). Because each conversion of the "currency baseline set" was independent of any (publically) announced program, the "money handling industry" had to scramble quite a bit each time; because there was no widespread "beta testing" of the new bills, many sites using bill-recognition devices (like self-service checkouts) did not accept "new" bills for weeks after their rollout (until firmware was upgraded).

Needless to say, the currency-handling industry managed these changes in the USA more-or-less without any major problems. By comparison, just think of the "fun" in the telecom industry if, say, SS7 had a "significant" change every two or three years...

HorsePockey. Digital smidgital..

Let me relate a story I was told. A engineer was tasked to retrain skilled mainframe COBOL programmers to the modern world. His part was to teach them assembler/ some machine architecture; so they'd have some clue as to what what going on Within the Beast. Later would come C, Java, etc.

As some point he was explaining CPU registers, and how it was faster to have a variable in one vs fetching it from elsewhere. Whereby, one of the smarter students asked "Why don't you just declare more registers..?" and he realized they did not grasp the basics.

He stopped, and explained a register was a special physical memory location, made up of IC's [well, cells of same] capacitors, etc. If you wanted more, you designed a new CPU chip. He said the looks on the class faces was priceless and almost scary -- never before had these skilled people grasped how things worked inside the box.

The parallel? Mr. Spade seems to think he can declare more digits in the switch. "Make it so.." said Capt. Picard.

Ain't so. The switch is HARDWARE, the 10-digit long registers are, like every part, highly optimized for speed, reliability, and low loading. It has man-years of engineering and code and testing and upgrades to keep it going. You start forklift upgrading parts and stand back...

(And that is ONLY Ma's switchers. You also need to replace every dialer program that stores numbers, & you name it on the customer's premises.)

My employer at the time had a Rolm PBX and had a nightmare getting the update done properly. For over a year, folks at one office were having to call the operator to place calls to the new area codes.

--scott

It wouldn't be the first time.

Just like iPhones. ;-)

The hard-wired switches are gone from the U.S. and (for the most-part) Canada.

How long have we had stored program controlled end office switches now? They became common by 1980. And, they enabled subscriber dialing of international numbers of varying length with delimiting by timeout or DTMF "#"

The COBOL programmers understood just fine. Back on computers like the IBM 1401 and the low-end IBM 360 series, all of the programmer visible registers were in fact stored in core, so declaring some more variables in storage was just as fast. It's not their fault that the hot-shot chip designers figured out a way to make registers faster than main storage.

Helpfully, John

PS: And watch those young whippersnappers just try to translate a simple COBOL picture like Z,ZZ9.99DB into Java.

Interesting point -- I wonder if that did indeed play a role?

Did the change to in-wats portability require every end office be modified or was that required only at the tandems?

That's a surprisingly complex subject: 800 portability was implemented to allow In-WATS subscribers to change their Inter-Exchange Carrier, so it has to be done at either the originating office or at a "LATA Access" tandem.

Before the call can be routed to the correct IEC, either the originating office or the tandem must "dip" the 800 database to determine the IEC, and then hand off the call to the carrier. It is the _IEC_ which determines the "POTS" number to which the traffic will be delivered, and most of them guard that information like a jewel, lest their competitors gain valuable business intelligence about time-of-day routings, call center load factors, etc.

Long story short: some offices have to hand off 800 calls to Access Tandems, some don't. It's not just a routing issue: many WATS numbers serve both "Band 9" (Intra-LATA) and Inter-LATA traffic, so the ILEC/CLEC could route many calls directly to the destination, but (as I said), the IEC's are very reluctant to give away the terminating numbers, and thus they often demand that _all_ traffic be routed to them, even if it's Intra-LATA.

Bill Horne Temporary Moderator

Thing is that right now it's purely political. I recall reading a simplified "Add 9" that someone on c.d.t. had brought up though I can't recall who.

You could simply roll that out and do a massive ad blitz reminding people to dial the additional digit.

Most of the software already supports international numbering plans. Even small switches like the Definity series have European ARS plans.

Are you talking about Y2k changes or the explosion in new area codes and exchanges that occured at that time?

Many private PBX vendors were unprepared, in a variety of ways, to properly track rapid new code assignments and get their PBX tables properly updated.

No hardware you say? How about the zillion little point of sale terminals at gas pumps, etc.

If anything the new currency is simpler to detect denomination than the old. Shine a UV light on it and look for the color and position of the return reflection.