[rfc-i] Is there a use case for 2119 keyword markup?

Nico Williams nico at cryptonector.com
Thu Jun 19 12:56:06 PDT 2014


On Thu, Jun 19, 2014 at 2:41 PM, John Levine <johnl at taugh.com> wrote:
> I can promise that if you do that, the first time MUST appears in
> unmarked upper case, we'll have a meta-argument about whether someone
> forgot to add the markup.

It's like so many nits: we call them out as such and that's the end of
it.  And frankly, we could stop calling out the nits -- that's what we
have an RFC-Editor for.

Anyways, that the nitpickers among us might have more occasions to
pick nits is not much of an argument against semantic markup.
Stronger arguments have been made _for_ semantic markup.

FWIW, I'm sold on semantic markup.  I especially like the <must/> (and
so on) approach: because it's proper markup that will survive passes
through XML processors (unlike entities), and because its short
enough.  I do also like &must;, but only if it maps to something like
<must/>.

I would also like an attribute or tag for indicating that some text is
INFORMATIVE.

Nico
--


More information about the rfc-interest mailing list