RE: : Diameter SIP Application AVP and command codes

Hi,

> if no effort is done for the alignment, the two specifications will
> have a lot of differences at protocol level, while being almost
> identical at a concept level. That means that most developers will
> find themselves building two protocol interfaces for their Diameter
> SIP Authentication nodes.
> 
> In my personal opinion, there should be an effort in making these
> two specifications compatible.

I agree.


BR,
Mikko


> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]]On Behalf Of
> ext German Blanco (E2/EEM)
> Sent: 10 August, 2005 12:00
> To: 'Warren, Dan, VF UK - Technology (TS)'; Loughney John
> (Nokia-NRC/Helsinki); Garcia Miguel.An (Nokia-NRC/Helsinki)
> Cc: [email protected]
> Subject: RE: [AAA-WG]: Diameter SIP Application AVP and command codes
> 
> 
> Dan,
> 
> I would be also of that 'old school', only problem is I wasn't there
> when all this started :-)
> 
> All (I hope somebody else apart from the people that are/were in 
> 3GPP reads this),
> 
> could anybody tell me if there was any discussion at all about this
> issue in the last IETF meeting?
> 
> I understand that the initial goal of this draft (please correct me 
> if this is wrong) was to have a *superset* of the Cx application.  
> That is, to have a Diameter application in IETF that allowed several 
> methods for authentication of a SIP user, one of which could be Cx.
> To me that makes perfect sense.
> 
> As it is, and as Dan said, there is a GREAT deal of common things, 
> and that is what I call a "double specification" (the same thing 
> defined in two places).  Not only that, if no effort is done for 
> the alignment, the two specifications will have a lot of differences 
> at protocol level, while being almost identical at a concept level.
> That means that most developers will find themselves building two
> protocol interfaces for their Diameter SIP Authentication nodes.
> 
> In my personal opinion, there should be an effort in making these
> two specifications compatible.  Could it be possible to move in the
> direction of removing common things and reusing existing codes so
> that we end up with that "superset"?
> 
> Cheers,
> 
> German.