Re: [Dime] [ALU] WGLC #1 for draft-ietf-dime-rfc4006bis-02

[Dave Dolson <[e-mail]>]

Maryse,
Thanks for doing some research and pointing this out.

In RFC 6733, RFC4282 is used for two things:
1. to define "Network Access Identifier", for use as realm names, which are "piggybacked on the administration of the DNS namespace"
- so DNS restrictions would have to apply here.

2. Defining User-Name AVP, which is a NAI, but specifically "of type UTF8String ... in a format consistent with the NAI specification [RFC4282]"
- (see section 8.14 of RFC6733)
- so User-Name is defined to be the UTF8 subset of RFC4282.

So I claim that although RFC4282 is mentioned, RFC6733 intends that user names in Diameter be limited to UTF-8, hence compatible with RFC7542.


-Dave


-----Original Message-----
From: Gardella, Maryse (Nokia - FR/Nozay) [mailto:[e-mail]] 
Sent: Wednesday, May 3, 2017 10:37 AM
To: Dave Dolson; Alan DeKok
Cc: Yuval Lifshitz; jouni korhonen; [e-mail] list; [e-mail]
Subject: RE: [Dime] [ALU] WGLC #1 for draft-ietf-dime-rfc4006bis-02

My mistake, it should be RFC 6733
Maryse

-----Original Message-----
From: Dave Dolson [mailto:[e-mail]] 
Sent: mercredi 3 mai 2017 16:19
To: Gardella, Maryse (Nokia - FR/Nozay) <[e-mail]>; Alan DeKok <[e-mail]>
Cc: Yuval Lifshitz <[e-mail]>; jouni korhonen <[e-mail]>; [e-mail] list <[e-mail]>; [e-mail]
Subject: RE: [Dime] [ALU] WGLC #1 for draft-ietf-dime-rfc4006bis-02

RFC4282 is also obsolete, and RFC7542 explains the problems with it.
I don't think we should introduce RFC4282 at this point.

(And sorry, I don't see RFC 6377 referring to 4282)


-----Original Message-----
From: Gardella, Maryse (Nokia - FR/Nozay) [mailto:[e-mail]] 
Sent: Wednesday, May 3, 2017 9:21 AM
To: Alan DeKok; Dave Dolson
Cc: Yuval Lifshitz; jouni korhonen; [e-mail] list; [e-mail]
Subject: RE: [Dime] [ALU] WGLC #1 for draft-ietf-dime-rfc4006bis-02

Hello all,

For the new AVP, no question: RFC 7542 should be used.
I have not the full overview of 3GPP specs used for reference to NAI, and based on:
- assuming the TS 23.003 (Numbering, addressing and identification) is an important spec to consider, the RFC 4282 is used
- RFC 6377 DBP also referring to RFC 4282
 
I would tend to agree on at least using RFC 4282 as the reference for the END_USER_NAI in Subscription-Id-Type for RFC4006bis.
Whether to directly refer to RFC7542, I cannot confirm whether this is acceptable or not.  

BR
Maryse

-----Original Message-----
From: Alan DeKok [mailto:[e-mail]] 
Sent: mercredi 3 mai 2017 00:47
To: Dave Dolson <[e-mail]>
Cc: Yuval Lifshitz <[e-mail]>; Gardella, Maryse (Nokia - FR/Nozay) <[e-mail]>; jouni korhonen <[e-mail]>; [e-mail] list <[e-mail]>; [e-mail]
Subject: Re: [Dime] [ALU] WGLC #1 for draft-ietf-dime-rfc4006bis-02

On May 2, 2017, at 5:51 PM, Dave Dolson <[e-mail]> wrote:
> 
> Thanks Alan.
> Do I correctly hear you saying we should replace all references to RFC 2486 with RFC 7542?

  Yes.

  It's 2017.  Independent of RFC 7542, *inter-operable* implementations just have no business using non-UTF8 identifiers.

  Alan DeKok.