Re: Structure of tel: URI

Yes, and that's a problem that's going to be causing confusion in the 
future if the allocchio draft goes forward. We had, I believe, 
intentionally decided not to support the various pauses and such in 
2806bis, since they are dial strings, not directory or routing numbers. 
The idea was that adding pauses and such depended too strongly on the 
local environment to be useful in a URI.

Michael Hammer wrote:
> Also, note how phone-string is being defined in:
> 
> draft-allocchio-gstn-04.txt:
> 
>       phone-string = 1*( DTMF / pause / tonewait / written-sep )
> 
>       DTMF = ( DIGIT / "#" / "*" / "A" / "B" / "C" / "D" )
> 
>       written-sep = ( "-" / "." )
> 
>       pause = "p"
> 
>       tonewait = "w"
> 
>       gstn-phone = ( global-phone / local-phone )
> 
>       global-phone = "+" 1*( DIGIT / written-sep )
> 
>       local-phone =  [ exit-code ] dial-number
> 
>       local-phone =/ exit-code [ dial-number ]
> 
>       exit-code = phone-string
> 
>       dial-number = phone-string
> 
>       subaddr-string = phone-string
> 
>       post-dial = phone-string
> 
> 
> Mike
> 
> 
> 
> 
> At 10:22 AM 8/7/2002 +0100, Lawrence Conroy wrote:
> 
>> At 5:35 pm -0400 6/8/02, Michael Hammer wrote:
>>
>>> Going from memory, one time when I looked into this, there was a 
>>> difference between global and local numbers in their use of 
>>> over-decadic digits.  Local allowed them, global did not.  I would 
>>> have to chase down the reference.
>>>
>>> Is this saying that over-decadic can be used in global or that local 
>>> digits can now not be over-decadic.
>>>
>>> Mike
>>
>> Hi Mike, Folks,
>>   "Over-decadic" was used for local numbers only to express RNs that 
>> were (?are?) used within some Netz.
>> e.g. Cxxx and Fxxx to indicate "this is a network-internal routing 
>> number".
>> It's mentioned in the BNF (as HEXDIG) and in ss 5.1.2 (line 264 onwards).
>>
>> The suggestion was that, as routing numbers are "network-internal", 
>> there's no need for this to be
>> allowed in global numbers - an RN is not (necessarily) valid outside 
>> its context. Hence 5.1.2 says
>> that it's unspecified but not prohibited for local numbers only.
>>
>> James Yu's draft on RN and CIC may make use of HEXDIG, so please let's 
>> not prohibit it.
>>
>> all the best,
>>   Lawrence
>> -- 
>> -----------------------------------------------------------------------
>> Roke Manor Research    : This information is provided "as is" and is not
>> <mailto:[email protected]>: intended to create any contractual or legal
>> <tel:+441794833666>    : relationship.
>