TELECOM@Usc-Eclb.ARPA (04/07/83)
TELECOM AM Digest Friday, 8 April 1983 Volume 3 : Issue 22 Today's Topics: American Bell Speaker Phones Phone Rate Restructuring Headsets Vs. Headsets Centrex/FRS - Routing Table Games [The date of this digest is one day ahead (so was Yesterday's) I will adjust for this over the weekend. --JSol] ---------------------------------------------------------------------- Date: Wed, 6 Apr 1983 11:11 EST From: Chuck Weinstock <Weinstock%Tartan-20@CMU-CS-C> Subject: Speaker Phones [TELECOM Digest V3 #21] American Bell has a new "speaker phone" they call the Quorum. It is designed to be used in a conference room, and has a rather strange appearance, being a rod sticking straight up into the air (I assume this is the microphone). It apparently doesn't suffer from the problems of echo and voice lockout. Drawback: they want over $1,700 + $250 installation! Chuck ------------------------------ Date: 6 Apr 83 10:32 PST (Wednesday) From: Thompson.PA@PARC-MAXC.ARPA Subject: Re: TELECOM Digest V3 #21 cc: Thompson.PA@PARC-MAXC.ARPA More regarding the whole phone rate flap. I agree with Lauren that the whole system as proposed will not wash with the public. I, too, think that the current system being imposed will produce a different political, rather than regulatory solution in fairly short order. I can't imagine that the public is willing to throw away the concept of a phone for every home without a major flap. I was of the opinion, all along, that if it wasn't broke they shouldn't have tried to fix it. On long distance access charges. I am hard pressed to see the rationale that we should get stuck with flat rate for long distance access when they are no longer willing to give us flat rate for local access. If this is supposed to be true dereg then the whole thing should be moved to "cost of service". In that case, access to a long distance carrier is a half of a local call. The long distance carrier should then be billed a half measured charge at the other end and pay it to the terminating, far end local company. Geoff [I think you miss the point, it's not Flat Rate Long Distance, it's a charge to get on, and a metered charge to use the service. Perhaps that metered charge will be the same no matter where you call (eventually?), but imagine paying $7.00/mo + measured rates per call for local service, and $7.00/mo + measured rates per call for long distance. --JSol] ------------------------------ Date: 6 Apr 83 13:42:04 PST (Wednesday) From: Hamilton.ES@PARC-MAXC.ARPA Subject: Headsets cc: Hamilton.ES@PARC-MAXC.ARPA How come everybody and his brother isn't clamoring for headsets? This "pinch handset between ear and shoulder" business is for the birds. And these kludges that you clamp on a handset to rest on your shoulder are utterly worthless. Why aren't headsets the default? Anybody know where I can get a CHEAP, LIGHTWEIGHT headset with a modular jack that I can carry around with me? Especially with all these cordless portables coming out -- why are they all handsets instead of headsets? Seems like it sort of defeats half the purpose... --Bruce ------------------------------ Date: 7 Apr 1983 0333-EST From: Hobbit <AWalker@RUTGERS> Subject: Routing table games Interesting recent bug: Rutgers is on a centrex in a split ESS office, no less, and the software to keep it straight must be hairy. Lots of the phones have toll restriction patterns. One type, arbitrarily named a B phone, may call a limited out-of-area radius, about 100 miles worth. This includes distinctions within our own area code [201], in that you can't call places way up in North Jersey near the NY border, but you can call New York City, Philly, and spots in South Jersey. It is really a table built to include a roughly circular calling limit. Recently I noticed that the error return for an attempted out-of-area call was being handled wrong: Instead of giving a recording or reorder, the calls were being *passed* to a couple of exchanges in Union, NJ. Specifically, anything with an area code ending in 7 would map the next 4 numbers into the 687 exchange. NPAs ending in 6 wound up in 686. Therefore if you dialed 617-253-6062 you'd get 201-687-2536. It took a while to figure out that this was happening, because not all numbers dialed mapped to real defined numbers in those offices [were met with a really crufty crossbar reorder signal]. Finally I made a wild guess as to who was in what central office in New Brunswick, and hit it on the first shot: I got a nice friendly guy at the SCC who I explained the problem to [Repair was no help, because as soon as I said something about an unpure routing table they got *very* confused!]. He actually understood, after I gave him an example number at Rutgers that was doing this. He had a look at that extension's status bits and realized what was going on. He told me to call back in an hour and by that time he should have had it fixed. An hour later I duly called back. He explained things as follows: There exists a service called Flexible Route Selection, which is basically an optimizer. If you have a centrex with tie lines to X, and some WATS lines, and some other regular ones, FRS will figure out where your call is going and route it the cheapest possible way. The service costs a lot, and is only sensible for large business applications. The B phones have something similar to this service, apparently, which is how they worked out the restriction tables. If you dial a number that is in the ''more expensive'' table, you get routed into limbo which tells you that you can't call there. A small fix could let such calls grab a different trunk and be completed; a *bug* could let calls get misrouted to Union, NJ. The guy was really nice about it, first really human switchman I've ever talked to. He did indeed fix the problem; it took all of 5 minutes to install a small patch in the table. The problem with FRS, as I see it, is that it takes forever to do the table search to find out if you can make your call or not. When you dial a number that is near the edge of the calling area [where it therefore can't consult the *local* tables], even by ''confirming'' the call with the # key doesn't help the as much as 3-second delay before it gets out of the office. Well, crufty algorithms aside, apparently the demand for such a thing is enough for it to be implemented. Does anyone know more specifics about it? The switching type didn't really go into intense detail about how it worked or why; I've reproduced what he did say as best I can. _H* ------------------------------ End of TELECOM Digest ********************** -------