In article , Helen Deborah
Vecht writes
01532 and 01734 were valid dialling codes for several years

01532 was never valid. 01734 was.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article , Stephen Osborn
writes
'phONEday' was in 1995 and all STD codes that did not start 01 had a 1
inserted.

Except for the five that got completely changed.

That was Easter Saturday so there more time than usual to sort out any
problems, also the network load the following week would be lower than
normal.

Even so, it almost broke. Over a quarter of calls were misdialled on the
first day; 30% was the "the network will break" line.

Reading was changed to 01734 in 1995 as part of phONEday but that number was
already getting close to full and the change to 0118 was already planned.

Not so.

It was not implemented until c. a year later to let people get used to the
previous set of changes.

That would have been silly, given it wasn't done anywhere else. If it
was certain that Reading would be about to fill, it would have been
better to do it with the other five.

Nobody was quite sure whether Reading was going to fill up, or if
somewhere else would beat it, nor what the best long-term strategy was
with something like 30 areas approaching trouble. So 0118 was held in
reserve for the next place needing transition - this turned out to be
Reading.

IMHO a lot of the subsequent problems were caused by Oftel not really
knowing what they were doing.

That I *can* agree with.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article ,
Martin Underwood writes
By the way, how did changing from 0171 xxx yyyy or 0181 xxx yyyy to 020 7xxx
yyyy or 020 8xxx yyyy help alleviate the shortage of available numbers in
London? It didn't increase the number of available phone numbers

Actually, it did: it made the 70xx, 71xx, 80xx, and 81xx blocks
available.

In addition, the costs of advertising the changes were shared among
telcos in proportion to the number of number blocks they had allocated.
This caused a sudden rush of "oops, we don't seem to need this number
block after all" letters to Oftel, freeing up enough numbers to last
several years!

OK, so there's scope for additional
district codes beginning with digits other than 7 or 8, but it's not
districts that are in short supply, it's subscriber numbers (the xxxx in the
above example).

Actually, it was districts (as you call them) that were in short supply,
specifically in the 0171 area. If a location fills up a Director Code
(as they were called when I started on this stuff) then another code can
be allocated to it. The problem was that the "inner" area had used most
of the 799 codes available.

The renumbering would improve things by, in order:
- allowing use of 201 more 7xxx codes;
- allowing the use of 8xxx codes in the "inner" area;
- allowing the use of 2xxx-6xxx and 9xxx codes.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article , Stephen Osborn
writes
Actually the local exchange simply routes all numbers that start with a 0 to
the associated trunk exchange (properly called a DMSU, for Digital Main
Switching Unit).

False.

The RCU (if one is involved) routes all calls to the DLE. This then has
routeing tables which say whether to send the call to an attached RCU,
to a DLE over a junction, or to one of the parent "tandems".

For example, the DLE to which my RCU is connected has 11 RCUs connected
to it as well as some subscriber loops of its own, and serves numbers on
01223, 01284, 01440, 01799, and 01954. 01223 is served by 6 different
DLEs at two physical sites, and the other codes I've mentioned there
share them in various ways:
01223 A B C D N P
01284 A B D N P plus a sixth DLE elsewhere
01440 A C D P
01799 A B P
01954 B C D P

Meanwhile other codes like 01279, 01353, 01366, 01485, 01553, 01638,
01760, and 01842 connect to A, B, C, D, and N as well as to other DLEs
not in Cambridge.

Calls within between those DLEs will *not* be sent to a tandem, even
though they need a 0 to be dialled.

Oh, DMSUs have all been replaced by NGSs.

The DMSU does geographic mapping, routes the call to the relevant DMSU on
the other end which in turn routes it to the relevant local exchange.

http://www.davros.org/phones/btnetwork.thml explains this in more
detail.

If it is a non-geographic number (07*, 08*, 09*) the DMSU routes it to a
special platform that does really clever lookups.

Or routes it to another telco.

don't
forget that at some stage 9,999 lines have to be connected up to each local
exchange.

Actually, 4096 lines is becoming the standard unit with the switch to
21CN; before that, it would be 3000 to 4000.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article ,
Martin Underwood writes
What about the situation where the same code is used by several towns and
villages, each of which has a telephone exchange. My code is used by two
moderate-sized towns and many neighbouring villages. I know that my village
has its own exchange (the building is about 100 yards from me right now!).
Presumably some form of supernetting is used: the first one or two digits of
the subscriber's number determine which exchange (consolidation device) the
call is routed to.

Example: 01954.

21xxxx Madingley concentrator - Cambridge C DLE
23xxxx Swavesey concentrator - Cambridge C DLE
25xxxx Cottenham concentrator - Cambridge C DLE
260xxx Willingham concentrator - Cambridge B DLE
261xxx Willingham concentrator - Cambridge B DLE
262xxx Willingham concentrator - Cambridge B DLE
267xxx Elsworth concentrator - Cambridge B DLE
268xxx Elsworth concentrator - Cambridge B DLE
71xxxx Caxton concentrator - Cambridge D DLE
78xxxx Crafts Hill concentrator - Cambridge Central DLE

So from a 21xxxx number, all calls will flow up to Cambridge C. Calls to
other 21xxxx numbers return to Madingly, 23xxxx and 25xxxx go to other
concentrators, and all other "same dialling code" calls go over a
junction to another DLE (B, C, and D are in the same building, Central
isn't) and thence to the correct conc.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article ,
Martin Underwood writes
I didn't know that? So did they use the letter suffix to denote the year?

Initially, yes.

If
so, did it start at the same time as in Great Britain - ie A=1963, B=1964
etc?

Except only London used A.

If so, I presume it went out of sync in the early 80s when IOM used U
and GB used V.

Earlier: MAN xxx T was in use within days of S registrations appearing
in Great Britain.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article , John Rowland
writes
Instead of 2 x 10,000,000
numbers there are now100,000,000.
No, a significant proportion of those 100,000,000 are unusable, because they
start with 0, or 1, or 999

0171 had 7990000 allocatable numbers, from 200 0000 to 998 9999. 0181
ditto.

020 has 79900000 allocatable numbers, from 2000 0000 to 9989 9999.

[Numbers beginning 0 and 1 can be allocated in addition, but only for
certain special uses.]

.... also one leading digit (possibly 2) will
never be used, because that will be added to the beginning when the numbers
eventually become 020 abc def ghj.

No such plans (I really can't see London needing more than 80 million
phone *numbers*).

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article , Jack Taylor
writes
Just to add to John's reply, London numbers starting with 020 3xxx are
due to start being allocated this summer. Unlike 020 7xxx and 020 8xxx,
they will be assigned on a London-wide basis and will not be mapped to
any particular district within London.

Which, AIUI, was supposed to be the case with unallocated 7xxx and 8xxx
series numbers post-020 implementation

Correct. However, most telcos chose to keep the distinction.

When 3xxx numbers start being allocated, 7xxx and 8xxx will be closed
even though there are numbers left in them. This is to deliberately
ensure that 3xxx gets used in both "inner" and "outer" areas.

For those wondering why 3 has been chosen, the reason is that there is
no 01203 dialling code, while there are 01202, 01204, 01205, 01206, and
01209.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article , Richard
writes
We could have [2-9]xxx xxxx; to get the others we would have to
require people to dial the whole number for all calls (as you would
from a mobile),

Such numbers (e.g. 01242 19xxxx or 020 0xxx xxxx) are already allocated.
However, they should not be used where the number needs to be
advertised.

Perhaps Ofcom could do some advertising that actually works this time,
when London starts to get 3xxx xxxx numbers.

Advertising should start any day now. Whether it works will be another
question.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In message , at 09:01:17 on Mon, 3
Jan 2005, Clive D. W. Feather remarked:
http://www.davros.org/phones/btnetwork.thml explains this in more
detail.

http://www.davros.org/phones/btnetwork.html

works better.

--
Roland Perry