Proposed change to RFCXML: remove <author> from <section>

[paulehoffman]

Description

The current vocabulary allows <author> elements in <section>. This change was made sometime after the publication of RFC 7991. If I squint hard, I can imagine the idea that a section of a draft or RFC has an author that is not for the whole document, but the semantics of that seem dicey, and possibly will cause legal challenges. Please revert to the RFC 7991 semantics where <author> can only be part of <front>.

Code of Conduct

• I agree to follow the IETF's Code of Conduct

[jrlevine]

Unfortunately if you do that, it will break every published RFC. There is an "authors-addresses" section at the end where the prep tool copies the list of authors from the front. I believe Henrik did that as a quick hack, and here we are.

It would make more sense to add an <authorlist> element which can be the last element in <back> and put the authors in there, but then we're back the the issue we don't want to think about, making changes to the XML in published RFCs even if there are no semantic changes.

[paulehoffman]

Thanks, I had missed that. Yep, authorlist would have made much more sense.

We will soon be in a world where we can make changes to published RFCs that are not semantic changes. The addition of <authorlist> and removal of <author> from <section> should be considered. Otherwise, people who wrote this one section during working group discussion will want their names on that one section.

[jrlevine]

Otherwise, people who wrote this one section during working group discussion will want their names on that one section.

I take your point, but it's been almost six years and so far nobody's done that. Either way I agree it would be nice to fix it so the tags match the semantics.

[cabo]

If I squint hard, I can imagine the idea that a section of a draft or RFC has an author that is not for the whole document, but the semantics of that seem dicey,

That is not the semantics today.
The semantics today seem to be to make a paragraph with the author information given.

This (and any imagined usage restrictions) probably should be documented in the RFCXML Reference

For contributor sections etc., we use <contact for the same purpose.

[jrlevine]

That is not the semantics today.

That’s not the way the prep tool uses it, but the XML grammar lets you mix author and text elements. Let’s fix that before someone tries it.

[cabo]

That is not the semantics today.

That’s not the way the prep tool uses it, but the XML grammar lets you mix author and text elements. Let’s fix that before someone tries it.

We do that in contributor sections, just with the new <contact element which is the same as <author.

So far, nobody has proposed the semantics of including an author element to express authorship for that specific section. That idea took me by surprise.