rnv@motsj1.UUCP (Ron Voss) (10/12/88)
In article <telecom-v08i0150m09@vector.UUCP>, c3pe!wb8foz@decuac.dec.com (David Lesher) writes: > While in Frankfurt last week I noted that phone numbers varied in > length from 4 to 10 digits, maybe more. The longer ones seemed > to be DID into PBXs. > How does the CO know when it has all the digits? Does it time > out, or do prefixes carry implicit lengths? Prefixes carry implicit lengths, mostly, although numbers can be "data sensitive". However, there are no prefixes as we know them, althought they technically exist. The "Post" doesn't advertise prefixes, and doesn't list or give out numbers in that way. For example, a phone number may be written 12 34 56 or 123 456 (Germany generally uses " " instead of "-" as a delimiter), whatever seems to be easier to remember, or however a particular person has always done it. Technically, all numbers beginning with the "hidden" prefix (in my example, you can't tell what it is: 1? 12? 123?) will be six digits long. I've seen phone numbers within an area code of four to eight digits. Area codes are two to four digits long. Generally, the shorter your phone number, the smaller your town, and, the shorter your area code, the larger your town. They don't use the British trick of having the target area code dependent on the caller's location. Large companies are often given a "prefix" of usually four digits. Then they do what they want, like 1234 0 for the switchboard and 1234 5678 for extensions, where internally the extension is 5678, just like here. The net will, as always, I hope, correct any factual errors. -- Ron Voss, Motorola Microcomputer Div hplabs!motsj1!rnv CIS 73647,752 408-991-7390 Opinions: My own
peter@hpqtdla.HP.COM (Peter Locke) (10/19/88)
>How does the CO know when it has all the digits? Does it time >out, or do prefixes carry implicit lengths? Neither, really, for most of the European networks. The old mechanical networks operated step-by-step signalling. This means that as you dial, different exchanges in the network interpret the digits you are dialling to determine the route, they connect you through to the next exchange and ignores the dialling. The next exchange now continues this interpretation until the end exchange rings the phone and - usually - any further dialled digits are lost. In the UK, for PBX direct extension dialling, the PTT fits a 'balck-box' in the end exchange which 'stores' further digits and then passes them on when the PBX connects - I don't know how they tell how many digits are needed. For the new CCITT SS#7 networks, as found in most European countries to some degree ( and all of Germany ), a call can be either step-by-step or block. We are used to step-by-step and it's benefits to us as users ( the network can give busy or bad number indications immediately, it doesn't wait till the whole number is dialled in ) and the CCITT#7 protocol has a Subsequent-Address-Message defined to ease implementation. ANSI does *not* have the SAM, so I guess either you're 'stuck' with block. So, in summary, we do not need time-outs to know if the user has entered all the digits. Effectively, the network 'finds out' by trying it. And no they do not need to be fixed length for the same reason. Now, when our national PTTs have *serious* competition, I think this will all change as the astute might have realized that step-by-step ties up equipment for longer than block. With equal access, the trunk providers will not be happy with step-by-step. I'm holding my breath !