Re: [Dime] Benjamin Kaduk's Discuss on draft-ietf-dime-rfc4006bis-08: (with DISCUSS and COMMENT)

[Yuval Lifshitz <[e-mail]>]

 Hi Benjamin,You are correct, this is a copy&paste error from "Final-Unit-Indication" AVP (which may contain both types). If someone is using the newly introduced "QoS-Final-Unit-Indication" AVP, they should only be allowed to use the new "Redirect-Server-Extension" AVP in it.Should fix to:"If the Final-Unit-Action AVP is set to REDIRECT then the Redirect-Server-Extension AVPs MUST be present."
Yuval
    On Friday, August 10, 2018, 9:36:13 p.m. GMT+3, Benjamin Kaduk <[e-mail]> wrote:  
 
 Hi folks,

I finally got a chance to look at the -10; sorry about the delay.

Most of the changes look good, but I'm not sure I'm reading things right
with respect to my last DISCUSS point.  That is, in Section 8.64 (of -10),
we still see:

  The Redirect-Server-Extension AVP (AVP Code TBD13) is of type Grouped
  and contains the address information of the redirect server (e.g.,
  HTTP redirect server, SIP Server) with which the end user is to be
  connected when the account cannot cover the service cost.  It MUST be
  present inside the QoS-Final-Unit-Indication AVP when the Final-Unit-
  Action AVP is set to REDIRECT.

That is, Redirect-Server-Extension, specifically, MUST be present in
QoS-Final-Unit-Indication when Final-Unit-Action is REDIRECT.

Down in Section 8.68, we learn about QoS-Final-Unit-Indication:

  If the Final-Unit-Action AVP is set to REDIRECT at least one of the
  Redirect-Server and Redirect-Server-Extension AVPs MUST be present.

Maybe I'm just missing part of the nesting structure of the AVPs, but
aren't these two statements still in conflict (in that the latter allows
sending Redirect-Server without Redirect-Server-Extension, that contradicts
the former's MUST requirement)?

Thanks for helping me check this,

Benjamin