Re: [Dime] DOIC #54: OC-Report-Type as mandatory AVP

I thought our chairs had declared a rough consensus on this. Do I have that confused with another issue?

On Jul 16, 2014, at 7:51 AM, Steve Donovan <srdonovan@usdonovans.com> wrote:

> All,
> 
> Now that the restructure of the DOIC specification is complete we need to start resolving open issues.  Those that entered the issues are encouraged to work to get them closed.
> 
> I'll start on #54 in this email.
> 
> The question is whether the OC-Report-Type AVP is required any time that an OC-OLR AVP is sent.
> 
> The current version of the draft is the same as the -02 version, showing the AVP as being required.  This was based on an earlier attempt to close the issue prior to -02 being published.
> 
> Susan has stated that she does not think it should be required.  If I understand correctly this is based on the assumption that an HSS would never send an OLR with report-type of any value but host.
> 
> Susan, please correct me if I'm mistaken.
> 
> I don't believe this is a correct assumption.  It is reasonable for an HSS to send an OLR with report-type of realm.  This might happen in a non agent based deployment or in a deployment where the HSS (or more generically, server) has knowledge of the status of all servers for the application.
> 
> Based on this, I propose that the text be kept as is and that this issue be closed without changes to the specification.
> 
> The threshold for changing what is currently in the DOIC specification is rough consensus in the group that it should be changed.  Anyone with an opinion is encouraged to state it now. This is not a vote (we don't vote in the IETF...) but an attempt to see if we have consensus to either close the issue without a change in the specification or if we have consensus to change the specification.
> 
> Regards,
> 
> Steve
> 
> _______________________________________________
> DiME mailing list
> DiME@ietf.org
> https://www.ietf.org/mailman/listinfo/dime