Re: Review of draft-mohali-dispatch-cause-for-service-number-12

["Ben Campbell" <[e-mail]>]

On 19 Dec 2016, at 11:11, Cullen Jennings wrote:

>> On Dec 15, 2016, at 10:07 PM, Joel M. Halpern <[e-mail]> 
>> wrote:
>>
>> At this point I will defer to the relevant ADs.
>
> +1 on that :-)

:-)

>
>> As far as I can tell, although the entry was created by an 
>> Informational RFC, the registry still claims that it is standards 
>> track.
>> And since this document is defining behavior, it behaves more like a 
>> standards track document than an Informational one.
>>
>> But it is up to you folks.  In teh end, all I can do is raise the 
>> question, not decide it :-)
>
> So the registry takes PS to change it.

To reiterate a previous comment on the thread: This draft does not add 
an entry to the registry, rather it adds a reference to an existing 
entry. The only point of the registry change is to make it convenient 
for implementors to discover that this draft updates 4458, which 
registered the entry in the first place. I'm not convinced that's 
completely necessary. But it might make sense to relax the standards 
action for this particular entry for historical reasons.

(Recognizing that the SIP URI parameter registry is messed up, also for 
"historical reasons".)

> And by the current SIP rules, I suspect (not sure) that an update to 
> 4458 would also have to be PS. So really not sure how one gets around 
> this not being PS.

I think this is the important decision to make. Setting the draft status 
based on the registration policy is an exercise in dog-wagging.

>
>
>>
>> Yours,
>> Joel
>>
>> On 12/15/16 11:51 PM, Ben Campbell wrote:
>>>
>>>> On Dec 15, 2016, at 10:35 PM, Joel M. Halpern <[e-mail]> 
>>>> wrote:
>>>>
>>>> I see your point about this adding a value to the entry created by 
>>>> RFC 4458.
>>>> Is there a reason this can not simply be PS?  The fact that 4458 is 
>>>> Informational does not, as far as I can tell, justify continuing 
>>>> the error.  While this is for a 3GPP usage, it appears to have been 
>>>> reviewed by the Dispatch WG sufficiently to justify PS.
>>>> One could argue that there is a down-ref issue,
>>>> but the fact that the field is in a standards-track required 
>>>> registry would seem to make that a moot point.
>>>>
>>>
>>> Do you think it makes sense to make some new values for “cause” 
>>> into a proposed standard when “cause” itself is informational?  
>>> That seems like a pretty big downref issue, as such issues go. (For 
>>> the record, I could be convinced to re-run this LC as PS, but I 
>>> suspect that would lead to objections in the opposite direction.)
>>>
>>> Right now, the situation is a standards-action registry with a 
>>> informational entry. That’s clearly broken, but I’m not sure 
>>> that _this_ draft is the place to fix it.
>>>
>>> Also, if it makes any difference—even there there was discussion 
>>> in dispatch, this is not a dispatch work item.
>>>
>>>> Yours,
>>>> Joel
>>>>
>>>> PS: It would seem that WG discussion of that sort is something we 
>>>> would like to see in Shepherd writeups.
>>>
>>> There’s two paragraphs on the subject in section (1) of the 
>>> shepherd writeup :-)  (but it wasn’t a working group discussion 
>>> per se.)
>>>
>>> Thanks!
>>>
>>> Ben.
>>>
>>>>
>>>> On 12/15/16 11:28 PM, Ben Campbell wrote:
>>>>> Hi Joel,
>>>>>
>>>>> Thanks for the comments. There has been a fair amount of 
>>>>> discussion
>>>>> about the status of the draft. The situation is clearly not 
>>>>> optimal, and
>>>>> I welcome input on how to straighten it out.
>>>>>
>>>>> The rational so far has been that this draft updates RFC 4588, 
>>>>> which is
>>>>> informational. It adds some additional values and related 
>>>>> semantics for
>>>>> the "cause" parameter from 4588. It does not register new 
>>>>> parameters;
>>>>> rather it adds itself as a reference in the existing "cause"
>>>>> registration. This is mainly a courtesy to readers. (There is no
>>>>> sub-registry for "cause" parameter values.) We might could get by
>>>>> without that change, since in a perfect world people following the 
>>>>> IANA
>>>>> reference to 4588 would notice the "Updated by..." tag.
>>>>>
>>>>> The gotcha is that RFC 4588 would not be possible as an 
>>>>> informational
>>>>> today; it would not have standing to register the "cause" 
>>>>> parameter. But
>>>>> at the time it was published, there was a lack of clarity around 
>>>>> the
>>>>> "standards action" policy for the SIP URI parameters registry. 
>>>>> Making
>>>>> the new values from _this_ draft standards track, when the 
>>>>> parameter
>>>>> itself is not, doesn't seem appropriate. We had some discussion 
>>>>> about
>>>>> whether we should promote 4588 to PS, but there was not consensus 
>>>>> to do
>>>>> so when it was published, and I don't see reason to expect that to 
>>>>> have
>>>>> changed.
>>>>>
>>>>> This draft is primarily intended to meet a need in 3GPP, where I
>>>>> understand they are effectively already doing this. Would it help 
>>>>> to
>>>>> more tightly scope this as "Here's something 3GPP is doing..." 
>>>>> rather
>>>>> than as a general mechanism?
>>>>>
>>>>> Thanks!
>>>>>
>>>>> Ben.
>>>>>
>>>>> On 15 Dec 2016, at 21:57, Joel Halpern wrote:
>>>>>
>>>>>> Reviewer: Joel Halpern
>>>>>> Review result: Ready with Issues
>>>>>>
>>>>>> Major:
>>>>>>   This document defines a new code for use in SIP, and specifies 
>>>>>> new
>>>>>> behavior both for the code itself and for its use in 
>>>>>> history-info.  I
>>>>>> am thus confused as to how this can be an informational RFC.  It 
>>>>>> looks
>>>>>> like it either Proposed Standard or experimental.  Yes, I see 
>>>>>> that RFC
>>>>>> 4458, which this updates is Informational.  But just because we 
>>>>>> did it
>>>>>> wrong before does not make that behavior correct now.  In 
>>>>>> addition to
>>>>>> my understanding of the roles of different RFCs, I note that RFC 
>>>>>> 3969
>>>>>> and the IANA registry both state that this assignment must be 
>>>>>> made by
>>>>>> a standards track RFC.
>>>>>>
>>>>>> Minor:
>>>>>>  Given our emphasis on IPv6 over IPv4, would it not make sense 
>>>>>> for
>>>>>> the examples to use IPv6 addresses?  (Inspired by the Id-Nits 
>>>>>> alert.)
>>>
>>>
>>