Re: Question about BCP 14 / RFC 8174

[John C Klensin <[e-mail]>]

--On Tuesday, August 26, 2025 15:51 +0000 "Salz, Rich"
<rsalz=[e-mail]> wrote:

>   *
> Is there a rule somewhere that IETF documents (at least on the
> standards   *
> track) MUST reference BCP 14?
> 
> Nothing in the formal RFCs that document how the IETF works.
> There's an IESG statement:
> https://datatracker.ietf.org/doc/statement-iesg-statement-on-clarif
> ying-the-use-of-bcp-14-key-words/

Let me go a half-step further.  There is nothing, either in BCP 14
nor that IESG statement, that requires the use of BCP 14 vocabulary
or definitions.  The IESG statement even says "BCP 14 is an optional
tool, and not strictly necessary." At least in principle, nothing
would prevent a document from avoiding BCP 14 terminology or a
reference to it, then using the same words (even in capital letters),
and giving entirely different definitions for them or using different
words entirely.   Especially if the same words were used, I'd predict
that such a document would have a very tough time during Last Call
and IESG review as various people asked (demanded?) that the authors
demonstrate sufficient justification for the confusion that would
likely cause, insisted that the document be super-clear about what it
was doing (not just omitting the BCP 14 reference), and demanded that
the justification meet a very high bar.  But the main reason for not
doing it is that it would be confusing and generally dumb, not that
there is any specific rule saying that BCP 14 definitions MUST (sic)
be used.

    john