[rfc-i] RFC citations committee I-D issued

Joe Touch touch at isi.edu
Thu Feb 10 14:08:46 PST 2011



On 2/10/2011 11:23 AM, Brian E Carpenter wrote:
> Joe,
...
>> 2)---
>> Sec 4 seemed a bit odd to me. In specific, AFAICT:
>>
>> a) standards-track docs already can (and do) cite IDs non-normatively
>>
>>      this should be noted
>
> I agree that this is the case, but it appears to be forbidden by
> RFC 2026. In fact the normative/non-normative distinction was
> invented at some point without benefit of a BCP. We intended to
> deal with this pragmatically.

AOK, but it's definitely been allowed for a while.

>> b) it's not clear what kind of citation is now being allowed; this
>> should be made more clear
>>
>>      IMO, it is NEVER appropriate to cite an ID normatively
>
> Agreed.
>
>>
>> I.e., it's not at all clear to me what is now being allowed that hasn't
>> been before, and (if so), why it should change. IMO, IDs can/should be
>> cited for credit, but never for specific content detail, regardless of
>> track.
>
> Also for historical review, if they are the only source. I have a
> number of such citations in draft-hu-flow-label-cases; in fact that
> is what sparked my interest in this issue.

Not sure what "historical review" means. The doc needs to be assumed 
ephemeral. If there's something relevant, the bulk of the idea needs to 
be included in the main doc, i.e., not relying on the citation as a 
historical reference. Overall, the point is that IDs are not historical 
references; they should be cited (IMO) in the same spirit as email - to 
give credit.

> What's new, I think, is that we're writing this down, and recommending
> that the full date should be included and that the full draft-name
> string may be included.

I agree with this; if there are multiples on a given day, it should be 
easy enough to explain which version is being cited (if substantively 
different) or just cite the date (if the core idea is invariant over 
that day).

>> 3)---
>>     7.  While many modern Internet-Drafts are available from an IETF
>>         Secretariat tools web site even after expiration, normally the
>> This should be updated as:
>>     7.  While many modern Internet-Drafts are currently available
>>         from an IETF Secretariat tools web site even after expiration,
>>         normally the
>> I.e, it's not clear that the tools site is a good thing to refer to here
>> in specific, except as a current example. (the location where that
>> archive exists may change)
>
> But it's explicitly *not* a Secretariat-maintained site.

Right - that's the reason I'm pointing this out. The tools site on this 
is not something we should expect to rely on.

Joe




More information about the rfc-interest mailing list