14 yo girl electrocuted trying to charge mobile in bath

[stillp]

once a connection was established, there was simply a direct 'wired' connection between electromagnetic devices at both ends, thereby not really offering any opportunity for 'handshaking' in the normal sense.

Telex used +80 and - 80 volts in the loop +80 as a space and -80 as mark ( IIRC) and a Carpentor polarised relay was used to regenerate the signal at the receiving end to drive the punch solenoids

Telex connections... If the connection was direct machine to machine then the transmitting station could monitor the current in the loop to verify that the connection was still good but there was no handshake on standard telex equipment. This did not work when the circuit went through a repeater / regenerator at the telex exchange(s) along the routing of the connection.

There were devices which could be added into the loop to create a check sum on the data received which could then be sent manually back to the transmitter as a "message". The operator there checked if it matched his check sum and if it did he sent a confirmation message. As I recalll these were not approved by the GPO

When Bernard?

 

[PBC_1966]

Telex used +80 and - 80 volts in the loop +80 as a space and -80 as mark ( IIRC)

Yes, +80 for space, - 80 for mark. The same +/-80V supplies were also used over the local loop for dial impulsing and supervision purposes (seizure to originate a call, cleardown to end a connection etc.).

Again AFAIAA, the 'Telex' system was introduced as a network (with exchanges etc.) essentially for connecting teleprinters to one another. I'm sure Paul will correct me if I'm wrong!

Yes, the G.P.O. Telex network was developed in the early-mid 1950's as a dedicated network to replace teleprinter operation over existing telephone circuits.

Re the electronics, while I mentioned it earlier as later Telex machines started incorporating such, remember that these machines still had to interwork with traditional teleprinters over the network using the established protocols, so as far as that side of things is concerned, there was no change. There would have been nothing to stop two compatible machines from implementing their own type of handshaking or error checking once connection was established, but that would not have been an inherent part of the network, and obviously would not have worked when one of those machines had to "talk" to a traditional teleprinter, such as the Creed 444 which the PO/BT were still using extensively right through the 1980's.

 

[bernardgreen]

If that was done in the original 'electromechanical' days, how on earth was the checksum calculated

I was never able to fnd that out. But never forget that relay technology was ( and still is ) able to do some very complex logical and computational processes. The designers of relay logic were extremely skilled engineers.

It was possible convert £ s d ( pounds shillings and pence ) into either pence or pounds and decimals of a pound in order for a standard decimal calculator to process the divisions and / or multiplications needed to create invoices and then reverse the process so the invoice result could be printed in £ s d

For young readers £1 = 20 shillings, one shilling = 12 pence, 1 pence = 4 farthings

 

[stillp]

1 pence = 4 farthings

One penny Bernard! Shame on you!

 

[bernardgreen]

I hang my head in shame.......

 

[JohnW2]

If that was done in the original 'electromechanical' days, how on earth was the checksum calculated

I was never able to fnd that out. But never forget that relay technology was ( and still is ) able to do some very complex logical and computational processes. The designers of relay logic were extremely skilled engineers.

That's very true, but it could be a major mission (I remember, in my youth, creating a 4-bit full adder with relays, and that was bad enough ) - and, from what you said, it sounds as if such a 'checksum calculator' would have to have existed at the location of every machine which implemented that system - something which sounds hardly practical with relay logic!

Kind Regards, John

 

[bernardgreen]

would have to have existed at the location of every machine which implemented that system

There were only a few organisations that thought it necessary to verify inter branch telex messages.

 

[stillp]

But never forget that relay technology was ( and still is ) able to do some very complex logical and computational processes. The designers of relay logic were extremely skilled engineers.

Even today, much PLC programming is done with (virtual) relay logic.

 

[bernardgreen]

Even today, much PLC programming is done with (virtual) relay logic.

Very true, a way to migrate the skills of design from relays real to relays virtual

BUT with a PLC you add a relay using a few key strokes, using real relays meand drilling hols to mount the relay, wiring up the contacts into the circuit.

Suck it and see is easy with PLC "relays", when it was a 15 minute job to add a relay one was far less inclined to suck it and see and more inclined to get t right BEFORE pcking up the tools.

 

[stillp]

Very true Bernard.

 

[JohnW2]

Suck it and see is easy with PLC "relays", when it was a 15 minute job to add a relay one was far less inclined to suck it and see and more inclined to get t right BEFORE pcking up the tools.

The same complacency has arisen in nearly all programming as technology has evolved ...

... my first encounter with programming arose when it was decided that it would be a good thing for us to dip our toes into it at uni. Once a week, we spent a whole afternoon trying to write little bits of FORTRAN code on coding sheets. At the end of the afternoon, we handed our forms to a person who when away and transferred it all to punched cards. The cards were then shipped a mile or two down the road to "the university computer". Come the next week, we would each be given a bit of paper representing 'the output' from our programme.

At least initially, this 'output' usually consisted of a single statement such as "Syntax error in line 3" We therefore soon learned to try to 'get it right first time' but, even so, it took most of a term for us to satisfactorily produce a programme to calculate the Standard Deviation of a small handful of numbers!

Kind Regards, John

 

[stillp]

Very similar to my experience, trying to perform CAD in Fortran, via punchcards and a mainframe in City Hall, 4 miles from my workplace. At least I could cut and verify the punchcards locally.

 

[bernardgreen]

Programming EPROMS for micro-processors,

1984 programming took about 3 minutes, erasing the dufff program out of the chip with UV was about 15 minutes to ensure it was cleared. After a while the legs snapped of the chips

2016 Erase and flash programming of the EPROM inside the processor is less than a econd for a couple of K of program code. Suck and see is taking over,

 

[bernardgreen]

Very similar to my experience, trying to perform CAD in Fortran,

1971 in Dusseldorf. The circuit diagram was encoded in some format and sent to a university ( or similar ) in London where the PCB layout was produced on a fast computor and the results sent back about a week later. Red and blue tape ups were just as fast.....after a bue and red pencil design

 

[stillp]

Also because memory is so cheap, programmers don't need to write efficient code. Many have never used machine code - there's no need. Remember when we would spend an hour or so typing hexadecimal then when we tried to run the programme, nothing happened. At least John's "Syntax error in line 3" gave a bit of a clue! I wrote some instrumentation code for a major car manufacturer, which fitted into spare memory blocks in their engine management system. They spent ages looking to see where I'd added extra memory!