RE: Review of IETF netann Draft

Just for review: The netann draft suggests that multifunction media servers have common user names associated with the device.  Those common user names are services at the device.

As is stated in the draft, these user names are not user names *in the domain*, but are user names *at the device*.  There are no human users, other than the ones defined in the draft, at the media server device.  How (really, "if") they are addressed in the domain is beyond the scope of the draft.

Even though there will never be a name collision at the device, people have expressed concern (confusion?) about the user names being domain-wide.  In that case, name collision is possible.

The suggestions on the URI list (here) is to prepend or postfix something to the suggested user names.

What would people think about prepending "__sip-" to the user name.

This would result in a different set of user names:

  OLD         NEW
 annc       __sip-annc
 dialog     __sip-dialog
 conf       __sip-conf

Would that be obfuscating enough to ensure a non-collision?

> -----Original Message-----
> From: Martin Duerst [mailto:duerst@w3.org]
> Sent: Friday, February 27, 2004 4:51 PM
> To: Eric Burger
> Cc: uri@w3.org
> Subject: RE: Review of IETF netann Draft
> 
> 
> At 15:56 04/02/27 -0500, Eric Burger wrote:
> >If we were to assume that a URL -at a host- somehow leaked 
> out as a URI 
> >-in a domain-, would John's proposal, to prepend the user 
> name with 'sip' 
> >and postfix it with 'master' alleviate the potential problem?
> 
> Something in that direction, yes, very much. I don't care too
> much about the specifics. Maybe having prefixes and suffixes
> is a bit too complicated, so rather than sipanncmaster,
> better anncsipmaster or so, or something a bit simpler.
> 
> Regards,   Martin.
> 
> 
> > > -----Original Message-----
> > > From: Martin Duerst [mailto:duerst@w3.org]
> > > Sent: Friday, February 27, 2004 11:02 AM
> > > To: Eric Burger
> > > Cc: uri@w3.org
> > > Subject: RE: Review of IETF netann Draft
> > >
> > >
> > > At 17:24 04/02/23 -0500, Eric Burger wrote:
> > >
> > > > > From: Martin Duerst [mailto:duerst@w3.org]
> > >
> > > > > At 12:50 04/02/19 -0500, Eric Burger wrote:
> > > > >
> > > > > >The IETF Internet Draft Basic Network Media Services 
> with SIP,
> > > > >
> > > >http://www.ietf.org/internet-drafts/draft-burger-sipping-neta
> >nn-08.txt,
> > > > >amongst other things establishes a URI convention for
> > > > addressing named
> > > > >resources at an automaton (in this case, a media server).
> >
> >
> > > > One solution to this problem may be to change
> > > >      sip:annc@example.net....
> > > > to
> > > >      sip:special-annc@example.net
> > > > (choose whatever appropriate for the 'special' prefix).
> > >
> > >In theory, any special prefix suffers the same problem.  
> "annc" takes away
> > >from the namespace the same as "special-annc".
> > >
> > >In practice, it is in fact the way SIP is used that makes 
> the namespace
> > >issue a non-problem.  The "reserved" users are typically 
> at a device or
> > >proxy.  They are not exposed externally, as discussed in 
> Section 6 of the
> > >draft.
> >
> >I have way not enough SIP knowledge to try to give you examples
> >how this might change in the future. But I'm also sure that you
> >cannot predict the future. Better be careful now than sorry
> >later.
> >
> >
> >Regards,    Martin.
> 
> 
>

Received on Wednesday, 10 March 2004 13:15:52 UTC