RE: "*" and "#" signs in the userinfo of SIP URI

My problem here is that this effort is a couple of years
old, and we are
trying to finish the work.  There is no spec that I know of
that couldn't be
improved, this one included.  When do we stop adding good
new ideas?  The
codes really are Hex E and F, the pound (square) and star
are what the Bell
System engineers decided to mark on the keys.  I defined
them that way to
match the P and X and so the string would not need escaping.
 You can use #
and *, and you would have to escape in a sip/sips URI.

As I understand it, you want to delete the E and F, and, by
the way, leave
the P and X, which are not part of the 3966 syntax.

I want to leave it as is, which means you can have an
escaped * or # or you
can have (an unescaped) E or F.  If you don't want to use E
and F, don't.

And, Paul, I agree about the comparison problems plural.

Brian



> -----Original Message-----
> From: Paul Kyzivat [mailto:pkyzivatcisco.com]
> Sent: Thursday, March 01, 2007 10:11 AM
> To: Alexeitsev, D
> Cc: iptelietf.org
> Subject: Re: [Iptel] "*" and "#"
signs in the userinfo of SIP URI
> 
> I never liked the use of E and F for * and #. Its my
impression
> (somebody correct me if I am wrong) that of the 16
possible DTMF tones,
> 14 are mapped to 0..D, and the other two are mapped to
* and #. So there
> is at least a temptation to see this as a hex encoding
and call * and #
> E and F.
> 
> I'm happy to see * and # supported. I don't mind so
much if E and F are
> also supported, though it could present issues during
URI comparison.
> But dial strings have other issues in comparison, so
this probably is
> not a big deal.
> 
> 	Paul
> 
> Alexeitsev, D wrote:
> > The latest draft
draft-rosen-iptel-dialstring-06.txt defines a way of
> > transporting the "*" and "#"
signs using the "E and "F" signs
> respectively.
> >
> > "A new alternative value for the
"userinfo" parameter of the 'sip:' or
> > 'sips:' URI schemes is defined,
"dialstring". This value may be used in
> > a 'sip:' or 'sips:' URI when the user part is a
dial string. The
> > dialstring is a sequence of the characters 0-9,
A-F, P, X, '*' and "#'.
> > E represents *, F represents #, P is a pause
(short wait, like a comma
> > in a modem string) and X represents "wait for
call completion"
> >
> > What is the rationale behind this alternative?
> >
> > There is already a defined way of transporting
this signs and bringing
> > new alternative does support interoperability.
> >
> > RFC 3261 says:
> >
> > "The BNF for telephone-subscriber can be
found in RFC 2806 [9].
> > Note, however, that any characters allowed there
that are not allowed in
> > the user part of the SIP URI MUST be
escaped."
> >
> > The 3966 (obsolets 2806) defines the transport of
the hex digits and "*"
> > "#" signs:
> >
> > local-number-digits = *phonedigit-hex (HEXDIG /
"*" / "#")*phonedigit-
> hex
> >
> > Greetings,
> > Denis Alexeitsev
> >
> >
> >
------------------------------------------------------------
------------
> >
> > _______________________________________________
> > Iptel mailing list
> > Iptelietf.org
> > https://
www1.ietf.org/mailman/listinfo/iptel
> 
> _______________________________________________
> Iptel mailing list
> Iptelietf.org
> https://
www1.ietf.org/mailman/listinfo/iptel


_______________________________________________
Iptel mailing list
Iptelietf.org
https://
www1.ietf.org/mailman/listinfo/iptel