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

Phillip Hallam-Baker phill at hallambaker.com
Thu Jun 19 05:21:15 PDT 2014


On Wed, Jun 18, 2014 at 9:46 PM, Ted Lemon <mellon at fugue.com> wrote:

> On Jun 18, 2014, at 9:06 PM, Brian E Carpenter <
> brian.e.carpenter at gmail.com> wrote:
> > I don't think that's the reason. I think the reason is precisely to
> > avoid any argument about which words are magic^H^H^H^H^Hnormative
> > and which are plain English.
>
> But it won't do that, because whether or not you see the words rendered in
> a way that makes that distinction clear depends on whether they were
> correctly marked up and what presentation you are using.   E.g., if you use
> text, you will not see the markup.   Encoding normative meaning in the CSS
> seems like a really bad idea to me.


Here is the markup I use in the input format:

<p>Normative words MUST be rendered in capitals</p>

Here is how it might appear in (one of) the output formats:

<t>Normative words <wev>MUST</wev> be rendered in capitals</t>

...

<section title="Normative Requirements">

<section title="MUST">

<p>1. Normative words MUST be rendered in capitals</p>
...
</section>
</section>

If the text is nested in a span section then the span section is taken as
the scope of text to extract for the compliance language. Otherwise the
enclosing sentence is extracted.

The text TBS is also recognized and used to create a list of issues to be
updated.


> I also think the proposal Heather included is the wrong way to do
> > this. MUST is really a separate entity and should be treated as such.
> > Hence, we should predefine the relevant entities as &must; &should;
> > etc. That separates the issue of meaning from the issue of presentation.
>
> But authors won't be required to submit this stuff, and so the RFC editors
> will have to scour the documents with some kind of pattern matching
> algorithm and hope that they identify all the normative uses of these
> words, and only the normative uses.


Or they could use my tool which accepts HTML or XML2RFC format as input. It
runs on Mac, Linux and Windows.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.rfc-editor.org/pipermail/rfc-interest/attachments/20140619/e402e913/attachment.html>


More information about the rfc-interest mailing list