RSE Wiki Archive

All auto-generated text that appears in the other formats needs to also exist in the canonical XML file as well. That text needs its own markup. The auto-generated text includes:

• Section numbers
• Figure numbers
• Table numbers
• IPR boilerplate

Extend the concept of language tagging to at least examples and contact information to address potential CJK font confusion. No objections to this for v3 from rfc-interest. -00

Include feedback information so that generated documents can provide usable feedback links. Wondering how this is supposed to work: is it only for metadata, or is a processor supposed to turn it into text. If the latter, where in the RFC does the text appear? Need to take back to rfc-interest.

Make the <date> element optional; all of its content is optional already. No objections to this for v3 from rfc-interest. -00

There should be a way to mark a section as something that will not appear in the published RFC. No objections to this for v3 from rfc-interest. -00

Add a mechanism for a ASCII fallback for non-ASCII names and addresses. Both forms should be searchable. General agreement on this on rfc-interest. -03

Get rid of the restrictions in <postal>, and make it free text in order to better handle different styles of postal addressing throughout the world. General agreement on this on rfc-interest. -00

Clean up the set of overlapping attributes between <figure> and <artwork>.

Remove anything that has to do with horizontal or vertical whitespace.

Extend <figure> to support different types of artwork. General agreement on this on rfc-interest. -01

Suggestion that we allow figures to have text wrapped around them. General dislike for this, or thinking that we don't need it, on rfc-interest.

Extend <xref> so that subsection and/or anchors can be specified.

Remove the “pageno” attribute.

Allow multiple paragraphs in a list item; eliminate the need to use <vspace>. General approval of this idea from rfc-interest for v3. -00

Create a new list style for dictionaries. General agreement on rfc-interest. -02

Items that need to be discussed. Short discussion on rfc-interest starting 2014-01-09, no conclusion

1. Should the reference value reflect the style, and in that case what is the correct reference value to use for each of the styles.
2. Should the reference value be built from the chain of list elements?
3. Should it be possible to generate a reference that is a partial list of the chain of elements
4. Should the default string be something other than Paragraph XXXXX

There have been a number of times where I have had a style=“numbers” list and then in the main text referred back to one or more steps. By doing this using anchors rather than by hand it makes sure that the step number is correct.

Allow overriding the “anchor” attribute of an included <reference> element. No objections to this for v3 from rfc-interest. However, this seems unlikely to work without Processor Instructions, which are being deprecated.

Add a way to add prose to a reference that avoids abuse of <seriesInfo>. No objections to this for v3 from rfc-interest.

Allow <reference>s that identify a document set such as a multi-RFC BCP or a multi-document standard from another SDO. Lots of discussion on how to do this for v3 on rfc-interest. -02

Deprecate or remove the <format> element. One idea was to use new elements such as @rel; another is to just point to the main RFC Info page. Lots of discussion on this for v3 on rfc-interest.

In v2, there are a few elements that require their sub-elements to be in a certain order even though that order isn't really needed by the XML processor. For example, <address> requires the enclosed <postal>, <phone>, <facsimile>, <email>, and <uri> elements be in exactly that order. Given that each sub-element has its own name, the order should not be important. A few questions about how the current processors work, but no objections to this for v3 from rfc-interest. -00

<spanx> has both a weird whitespace model (“preserve”) and problematic styling. Deprecate it, and in its place add <tt>, <b>, and <i>. Not yet discussed on rfc-interest, but included in -00.

The <section> element could have an attribute to indicate that the section is not numbered. This is useful for top-level sections at the end of a part (such as “Contributors”), and also for low-level sections whose numbers would make the headings hard to read. Processors need to check that this attribute is used in allowed places. An alternative model would be to allow the <note> element (which is only allowed as part of <abstract> in v2) to appear in any of the <middle> and <back> parts as well; again, the processor would need to check the position of this element. No objections to this for v3 from rfc-interest. -00

Maybe allow for xrefs to be used in the <ttcol> element. It might also be reasonable to allow for iref and cref and eref and spanx to be used here as well. No objections to this for v3 from rfc-interest. -00

Add a <blockquote> element. This is to be used for real quotes, not just for text that is indented from both left and right margins; there should be an ability to do the latter as well. Added in -03