Re: [Gendispatch] Some data about Updates in recent RFCs (was Re: New Version Notification for draft-kuehlewind-update-tag-04.txt)

[Brian E Carpenter <[e-mail]>]

Below...
On 20-Jul-21 21:22, Mirja Kuehlewind wrote:
> Please see inline.
> 
>> On 19. Jul 2021, at 22:39, Brian E Carpenter <[e-mail]> wrote:
>>
>> On 19-Jul-21 20:14, Mirja Kuehlewind wrote:
>>> A citation gives you a reference from the new RFC to an existing one. 
If you also want to add a pointer from an existing RFC to a new one, you need to add meta data.
>>
>> Right, but I don't think the current text in the draft makes it clear when you would want to do this. Consider that "cited by" is already available in the tracker, e.g. https://datatracker.ietf.org/doc/rfc8200/referencedby/
>>
>> I chose that example to make it clear that "See also" is not the same as "Cited by". Which of the RFCs that cite RFC8200 would deserve a "See also”?
> 
> Thanks for bring this example up. This list at least shows that “cited by” is not useful in this context. I guess most of the documents on this list “use” RFC8200. So for the reader of the new documents, it’s important to have the pointer to FC8200 but 
any reader of RFC8200 would probably not really be interested in most of those documents on this list. I don’t think I have a perfect answer which of these documents could be “see also” (and I didn’t try to put in the time just now to figure it out) but I would expect a much shorter list.
> 
>> And, by the way, who decides, since this will be added after publication?
> 
> I have an open issue for this on github:
> 
> https://github.com/mirjak/draft-kuehlewind-update-tag/issues/7
> 
> However, I on purpose did not try to address it in the last revision. I 
believe having these tags is just part of the consensus process as the rest of then document as well. As such similarly as for every other issue that is raised by the IESG, the IESG can block the document until the issue is addressed. Not sure we really need to say anything explicitly in the 
draft?

"See also" is a bit different because it doesn't relate to normative effects in the way that "Amends" and "Extends" do. So it could affect any stream, not just IESG-approved documents. I guess that means that the RFC Editor is responsible. The old "Also RFCxxxx" designation that you find in the RFC index, starting with RFC29 and continuing at least through RFC2389, was certainly an RFC Editor decision. Possibly all of those should be changed to "See also"?

   Brian

> 
> Mirja
> 
> 
>>
>>   Brian
>>
>>>
>>> I don’t expect this will be used extensively as I do believe “amends" and “extents” covers the most important cases but it can still be useful (e.g. to connect a suite of recommendation documents…?).
>>>
>>>
>>>> On 19. Jul 2021, at 06:54, Brian E Carpenter <[e-mail]> wrote:
>>>>
>>>> On 19-Jul-21 16:14, Martin Thomson wrote:
>>>>> On Mon, Jul 19, 2021, at 14:10, Suresh Krishnan wrote:
>>>>>>> That doesn't surprise me. I think it's a useful relationship; for 
example 
>>>>>>> a large number of the RFCs cited in "IPv6 node requirements" could have a 
>>>>>>> "see also" relationship with each other, but it's quite different 
from "updates".
>>>>>>
>>>>>> Yep. Exactly. Fully agree,
>>>>>
>>>>> Isn't that what citations are for?  Keeping metadata lean would be most consistent with its prominence.
>>>>
>>>> Yes, in many cases a citation (and the implied "cited by") is enough, but "see also" is only intended for occasional use:
>>>> https://www.ietf.org/archive/id/draft-kuehlewind-update-tag-04.html#section-3-4
>>>>
>>>>  Brian
>>>>
>>>> -- 
>>>> Gendispatch mailing list
>>>> [e-mail]
>>>> https://www.ietf.org/mailman/listinfo/gendispatch
>>>>
>>>
>>
>> -- 
>> Gendispatch mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/gendispatch
>