There seem to be a few people interested in continuing the technology theme that has come up in Digby's thread on the latest telephone scam, so I propose to continue it here in order not to put off non-technical people who may prefer to read about the scam details.

Absolutely correct - I was employed to design telecommunications computing equipment and was surrounded by uniselectors, two-motion selectors, motor uniselectors, crossbar switches, reed relays and ultimately digital switches - and had to interface my designs with all of them at some time or another . Although we sold a lot of our Strowger equipment new to other countries, I am not sure what BT did with their old stuff, bearing in mind that it was severely worn out by the time they uninstalled it.

Yes, "Erlangs" - fortunately for me we had a maths wizard who kept me straight on that topic They don't work for statistical multiplexing as used in packet switching networks either. But the thing that always gets me is when you go into a public toilet and find something like 6 urinals, two handbasins and one very weak hand dryer. Since you spend longer at the dryer than either of the other two, Erlangs formula would tell you that you need more driers not fewer! It does seem a crazy requirement today, but if you go back to when the GPO were first putting in their automatic telephone system, for simplification of the metering they decided to charge per call, not by time. That much is for the benefit of the service provider. But by giving control of the call to the initiator they avoided putting themselves at risk from repeated complaints that the caller had paid for a call which had been prematurely terminated by the person not paying for the call. Mind you, how they expected the called person to actually listen and talk if they did not want to is another question .

With regard to your 1 - it is the case in many countries that either party can release, but in this country that is not consistent with past practice, and which rightly or wrongly BT have continued with. In any case, there would have to be a finite, albeit small timeout to guard against calls dropping out due to momentary line disconnections. These were naturally very common in the days of Strowger switches, relays, and open drop wires, but nowadays they are usually limited to those caused by BT Outreach poking about inside cabinets! You are right about your hypothetical criminal act (I know you wouldn't really commit one ) but in reality you only have to find the drop wire outside the house and cut it, and then you have unlimited time in which to commit it.

Regarding your 2 - in this country it always has been the initiator of the call who pays for it, except in the specific case of a Reverse Charge Call. (Apart that is when you are called on your mobile phone and you are roaming onto someone else's network, and then the called party has to pay to get the call from his home network to the foreign one. I am sure you are aware of this, though.) The choice is whether or not you allow the person paying for it to decide when to end the call. For historic reasons as explained, the UK network has always given this to the initiator, subject to the guard timeout.

I think some colleges must have bought some of it too for learning purposes.

Most of the work I did for telecomms carriers in the past was Voice/TDM/Frame (can someone fix this FECN thing!)/ATM on Newbridge/Alcatel/Lucent kit. Fun days! Customers never learning that you don't send ringing current into an E&M circuit was a common thing. I got my fair share of ringing current belts too... I considered it an initiation of fire

I'm not old enough to have worked on X25, but I did used to do work on some of the old legacy Timeplex kit which was fun

When I was at school, a good friend of mine's father happened to be some sort of director at BT and got us into Martlesham one weekend for some event. I remember being distinctly impressed with the stuff those R&D guys were working on!