[VoiceOps] [External] Re: 9-8-8 dialing when an outside line access code (9) is being used

Jay Hennigan jay at west.net
Mon Jul 18 02:17:47 EDT 2022

On 7/17/22 21:19, Hunter Fuller wrote:
> We operate a system with the "dial 9" scheme (apparently "useless"
> according to other posters - a truly insightful attitude that I love
> to see on this list),

It's not unusual in old-school PBXs and wireline POTS where digits are 
processed serially. With cell phones dialing is en-banc with a SEND 
button so digit patterns no longer need to be unique. See note on 
en-banc dialing below.

> so I can say that the expectation definitely is
> NOT for people to dial 9911. In fact, there is a whole law about it,
> which, like many, is written in blood:
> https://www.fcc.gov/news-events/podcast/personal-story-behind-karis-law

This did take some special programming, however. The leading 9 is a 
trunk access code and should return a second (often different sounding) 
dial tone. The law to which you refer is because of PBXs that weren't 
specially programmed to re-insert the stripped "9" and send the call on 
its way. The original expectation was "Seize an outside line (by dialing 
9) wait for dial tone, dial 9-1-1."

When I was programming Mitel PBXs back in the day, I ensured that both 
9-11 and 9-911 would get routed to 9-1-1 regardless.

> The difference is, if someone picks up a phone and dials 911, they
> want 911. They don't want an "outside line" so that they can dial a
> NANP 10-digit number beginning in 11, because no such number exists.
> The problem is, such numbers DO exist that begin with 88, so, we are
> in a bit of a pickle there. It seems the only solution is to do a
> timeout... yeesh. (Unless I'm missing something.)

Kinda, sorta. 7-digit local dialing is supposed to have been phased out, 
with all NANP numbers represented as 1+NPA-NXX-XXXX.

This means that after your trunk access 9, you should expect a 1 
(followed by ten digits for a regular phone number), a 0 for operator or 
011 international, or a three-digit code starting with 2 or 9 that until 
this week always ended in 11.

> Dialing 911 directly (not 9911, but just 911) has always worked here,
> long before Kari's Law, and it works without delay, as it should. I'd
> love to make 988 work the same way but I'm just not sure how to
> accomplish that.

Program 88 as a sequence to re-insert the stripped 9 and send 
immediately on trunks accessed by a 9, just like you do with 11.

If you still want to allow 7-digit dialing and have a local 88X prefix, 
or if your dialplan allows 10-digit calls without a leading 1, then yes, 
you'll need a timeout. Or make it 9-988 until they pass another law.

Note: In fact, en-banc cell dialing broke a few advertisements where a 
word was spelled longer than 7 digits. For example, "Dial 
1-800-HARDWARE" worked fine from a landline or (with prepended 9) from a 
PBX. As soon as the digits 1-800-427-3927 were dialed, the call would 
complete. With a cell phone, however, the number sent is 1-800-427-39273 
which doesn't match a valid number and the call would be rejected. Some 
cellular carriers have worked around the issue and truncate long strings 
to match the NANP.

Jay Hennigan - jay at west.net
Network Engineering - CCIE #7880
503 897-8550 - WB6RDV

More information about the VoiceOps mailing list