Installfests:Cats

Peter plp-ysDPMY98cNQDDBjDh4tngg at public.gmane.org
Fri May 13 19:32:21 UTC 2005


On Fri, 13 May 2005, Lennart Sorensen wrote:

> On Fri, May 13, 2005 at 09:07:47PM +0300, Peter wrote:
>> the telecom people who knew from experience that 000 is the number that
>> is least likely to be dialed by glitch in a rotary dialer system.
>
> Well glitches can dial 911 quite easily.  I had a phoneline a few years
> ago that went staticy on me, and sometimes just didn't work.  After a

If you have enough noise on a line then it will call ALL the possible 
numbers eventually. That was not what I was talking about. In a rotary 
system the 000 corresponds to the *longest* pulse sequence possible that 
is valid. If any glitches occur in the switching then they will mimick a 
*shorter* pulse sequence (as long as the delays and speeds are set right 
in the dialer and in the exchange - and they are so), or a *longer* 
pulse sequence. Exactly 0 (exactly 10 pulses) had the lowest probability 
to be jammed or glitched. 999 was likely a compromise adopted when 0/00 
started being used for international/interurban automatic dialling 
probably. Afaik 00 was used as the prefix for interurban dialing 
precisely for this reason (low probability of glitch = low number of 
complaints for undesired interurban connects, and low load on 
interurban trunk selectors).

111 is the number with the highest glitch probability in a rotary dialer 
system. I don't know how they got from this to 911. There must be some 
clever reason for that too. 911 combines the low glitch probability for 
the first number with fast dialing (000 is the number that is the 
slowest to dial on a rotary dial - probably not fast enough if someone 
is breaking into your house at the time). But there may be a different 
reason.

> Apparently a broken phone line shorting to ground can pulse 911 fairly
> regularly.  I suppose it pulsed other numbers too at random, but none of
> those ever called back, or at least not persistently enough to get
> through.

I'm a hardware guy and I can tell you that a faulty part or wire can 
send all sorts of interesting data if you wait long enough. No coding 
system will withstand that kind of 'brute force' attack for too long, 
and systems where this is important do detect noise and signal it in 
time (and some will even shut down the link is there is too much noise).

Peter
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml





More information about the Legacy mailing list