Re: Comments on RFC2806bis

Henning Schulzrinne <[email protected]> Tue, 21 May 2002 10:57:38 -0400
Newsgroups gmane.ietf.iptel
Organization Columbia University
Message-ID <[email protected]>
> - A node does not understand "m-rn" or "m-cic" and stops processing the SIP
> INVITE message
> 
> - A node does not understand "rn" or "cic" and routes based on the phone
> number
> 
> The choice is the latter.  This is because when the call arrives at a PSTN
> gateway or another SIP proxy server, that gateway/server may understand "rn"
> or "cic" or may be able to do another NP/800 database dip.
> 
> I suggest that 2806bis discuss whether a node drops or keeps "unknown"
> optional parameters when it receives the INVITE message and needs to send
> another INVITE message out.  It seems better to keep the unknown optional
> parameters in the message because the next node may be able to handle it. 
> 

2806bis is not SIP-specific, so it would not be appropriate to discuss 
transformations of URLs there. Generally, the normal URL rules in SIP apply.

> Can the country code contain hex digits? Is there the possibility 
> that country codes need more than three digits? Does East Timor have its 
> own country code?
> 
> [JY]
> According to E.164, country codes (CCs) can be assigned
> 
> - For geographical areas - CCs are one to three digits long.
> 
> - For global services -  CCs are always three digits long.
> 
> - For shared international networks - CCs are always three digits long.
> 
> In all three cases the CCs do not contain digits A thru F.
> 
> The global routing number I have in mind will always start with "+CC" where
> the CC does not contain digits A thru F and the digits after CC may contain
> hex digits.  So a global routing number, rn=+1-202-533-1234, would be the
> same as a local routing number, rn=202-533-1234;phone-context=+1. 
> 
> May be we can use "number-context" instead of "phone-context" so that it can
> be used for the phone number and routing number.

I thought somebody already complained when I had accidentally changed 
the label to "context". If a change is acceptable, I'd rather go for the 
shorter "context". In the end, I don't think the name much matters, as 
long as the description is clear.


> 
> James