[rfc-i] Summary: use case for 2119 markup

Nico Williams nico at cryptonector.com
Fri Jun 20 14:12:44 PDT 2014

On Fri, Jun 20, 2014 at 4:04 PM, Dave Crocker <dhc at dcrocker.net> wrote:
>> On Jun 20, 2014, at 4:24 PM, Andrew Sullivan <ajs at anvilwalrusden.com>
>> wrote:
>>> Wait a minute.  I thought we were _already_ accepting a process
>>> change, because the "canonical" version of the file is no longer to
>>> be the text file, but instead some other format.
> Yes, but... that's not carte blanche to mess with whatever whim hits the
> RSE (not that I'm saying they are whimsical, of course).

This thread is prima facie evidence that the RSE does not act as if
they had a green light to do what they please.

> Especially since the choice of canonical representation has been for one
> that is already established here, messing with document semantics
> crosses into new territory.

I don't a proposal to add markup for RFC2119 terms as changing
semantics.  Please explain how that perception is wrong.

RFC2119 doesn't mandate anything as to rendering styles anyways -- not
even as to case(!).

All that said, perhaps RFC2119 really does need an update:

 - to REQUIRE (self-referentially?) that terms it defines stand out
from other text when RFCs using them are rendered

 - to specify some typographic constraints on freedom of
representation of RFC2119 terms

Also, is there an RFC that requires the use of RFC2119 terms for
Standards-Track RFCs?  If so, what is it?  If not, should there be?


More information about the rfc-interest mailing list