Re: On XML and $EDITORs (Re: Things that used to be clear (was ...)) "Living Documents") side meeting at IETF105.)

[Phillip Hallam-Baker <phill@xxxxxxxxxxxxxxx>]

On Thu, Jul 11, 2019 at 11:09 AM Ted Lemon <mellon@xxxxxxxxx> wrote:

On Jul 11, 2019, at 10:25 AM, Julian Reschke <julian.reschke@xxxxxx> wrote:

The insertion of the second one could indeed be automated, but then you
would have to give a hint about normative vs informative.

That would actually be a really nice enhancement.   It could be the rule that if no reference says, it’s normative, if all references either don’t say or say normative, it’s normative, if all don’t say or say informative, it’s informative, and if some say normative and some say informative, it’s informative.   The idea being that some references to the same document may be normative and some informative, and it makes sense to be able to signal that in the XML.  Maybe every reference should say, but I’m sure that would be a pain.

It would also be nice if the publication date could default to “today” and if the draft number could be determined automatically based on what is most recent…

RFCTool does the following when converting documents.

1) References of the form <info="RFC822"> and <norm="RFC32"> are converted to the appropriate references. Same for drafts.

2) Defaults to today's date

3) works out the correct revision number so as soon as 

draft-hallambaker-mesh-architecture-09 appeared in the registry, my tools started generating architecture-10

4) MUST, MAY, SHOULD are recognized as normative language in the text and can be used to build an optional table of normative language. Alternatively, these could be rendered in bold.

5) Underlined text is rendered as fixed spaced font. This is because very few documents need underlining but most have references to identifiers.

What I don't bother to do is to implement all the curlicues on lists like starting at a particular item. This could be added but I have never found a need for it. And references to Word section headings seem to have broken recently... ooops.

The thing I would most like to add is the ability to identify words as the point at which defined terms are defined and used and use that to build an index. Easy enough to define the code to build the index, less easy to work out a convenient way to enter the information in a Word document.

I think it would be very helpful to agree on a set of markup tags. What I use currently is essentially re-purposing the xml2rfc tags. So this is the pre-amble in one of my files:

<series>draft-hallambaker-mesh-cryptography
<status>informational
<stream>independent
<ipr>trust200902
<author>Phillip Hallam-Baker
    <surname>Hallam-Baker
    <initials>P. M.
    <firstname>Phillip
    <email>phill@xxxxxxxxxxxxxxx
<also>http://mathmesh.com/Documents/draft-hallambaker-mesh-cryptography.html

The reason I need also is that my documents depend on features in the HTML output like superscripts and subscripts.

The only other XML style markup I use most of the time is to include figures and examples:

<figuresvg="../Images/Cogen.svg">Two party key pair generation.
<include=..\Examples\ExamplesAdvancedCoGeneration.md>

And yes, this isn't legit XML because legit xml is rather a bore. why type <a href="" when the tag always takes at least one attribute and it is usually the only attribute?

I could happily agree to move to {{a="http..."}}

There is an open source MIT license library for interacting with Word docx format and it works pretty well. I have not tried using OpenOffice but I strip out 90% of the markup as irrelevant in any case.