Source format, aka Submission format, Input format = the format used by authors, etc., when writing an I-D. Currently: XML (can be uploaded to the I-D submission tool currently), NROFF, .doc, .txt

• may or may not be the same as any other format (though it would make the workflow somewhat simpler for the RFC Editor if it were);
• will be converted to another format for further processing and publication if necessary

Canonical format = the authorized, recognized, accepted, official version of the document

• other document types may be derived from this format; may or may not be one of the possible output formats

Archival format = currently the printed, paper copy AND the digital .txt of the RFC

• currently the same as the canonical format

Output format, Display format = format as it may be read or printed after publication process has completed; currently ASCII, PDF, and HTML are common output or display formats

• For I-D: TXT*, PDF*, PS*, HTML.
• For RFC: TXT, .TXT.PDF, Enhanced PDF (rare)
• Outside RFC Editor: HTML (via rfcmarkup), HTML (via datatracker), PDF, perhaps others.

Pagination

• For: Ease of reference and clear printing; referring to section numbers is too coarse a method
• Against: Want a smooth reading experience regardless of page or screen size
• Additional thoughts: the RFC Editor does not recommend using page numbers as points of reference

Character encoding - ASCII

• For: Most easily searched and displayed across a variety of platforms. In extreme cases of having to retype/scan hard copies of documents (it has been required in the past) ASCII is significantly easier to work with for rescanning and retaining all of the original information
• Against: Too limiting with regards to internationalization issues

Character encoding - UTF-8

• For: Allows authors to spell their names correctly; certain special characters in equations or quoted from other texts allowed; citations of web pages using more international characters possible; in discussions of internationalization, actually being able to illustrate the issue is rather helpful, and you can't illustrate a Unicode code point with “U+nnnn”.
• Against: Exactly what characters are allowed and where the line should be drawn remains unclear (why some characters commonly used in European languages and not other, non-Latin characters? This is just pushing the problem around.)
• Additional thoughts: just moving from ASCII to UTF-8 (as opposed to UTF-8 and HTML or XML) leaves us with dependencies on the local file systems and processors to be configured properly and do the right thing with the document though adding a byte-order mark (BOM) to the beginning of each non-HTML UTF-8 text file will go a long way towards recognition; browsers will recognize UTF-8 and can declare the encoding definitively

Mobile Devices

• For: We should take their needs for format flexibility (reflow) in to account
• Against: Not enough people use mobile devices, and those that can can generally scroll, so this should be treated as an edge case at best
• Additional thoughts: interesting numbers regarding mobile device web browsing: https://vivipins.com/mobile-marketing-statistics/

Use of RFC-specific tools

• Against: We can't be that unique in our needs that we can't use commercial tools
• For: We have more control over the tools we write, and the audience that reads RFCs will always be capable of coding up something new if needed; we have xml2rfc to work from as a base and should perhaps consider how to retain nroff

ASCII art

• For: Dependence on advanced diagrams (or any diagrams) causes accessibility issues; forces people to rely more on words and clear written descriptions than the diagrams; each diagram is [relatively] simple and discrete
• Against: It does not allow for reflow; the limitations to the diagrams are a hindrance to visual thinkers
• Additional thoughts: If we go beyond ASCII art and have the normative diagrams be entirely separate documents, we may not need to limit ourselves to one graphic format (but doing so may make things simpler). Also, if we go beyond ASCII art, wneed to pick just one format: GIF? PNG? SVG?

Graphic art / Advanced diagrams

• For: makes it easy to include complex equations; some people consider graphic art easier to understand/read
• Against: may take pressure away from authors to focus on clear language
• Additional thoughts: if an ASCII format continues to be used, graphic art may be used and referenced as a separate document; it can be internal to a document in either the XML or HTML proposed formats

Equations

• For: Some authors have chosen not to publish RFC due to difficulty in displaying proper mathematical equations
• Against: So few RFC include mathematical equations that this should not be given any priority in the discussion of format

Metadata and tagging

• For: Ability to semantically tag some document info, at least authors' names and references is useful
• Against: Metadata is unnecessary overhead
• Additional thoughts: there is no complete list of tags that will be required for XML or HTML that would build-in required simplification and support for the archival nature of the series (that people can work longer with a simplified set of tags), and until we have that, we cannot talk about tags (note that a list has been started in the XML format I-D)

Containment

• For: Lack of containment for sections means that processing software cannot be fully aware of the document structure, and that is serious restriction
• Against: Containment is unnecessary and not compatible (or perhaps just not required?) with traditional HTML and word processor document
• Against: Requiring containment may limit the number of editors authors can use to create documents
• Against: Requiring containment would require every authoring format to be translatable to the submission format
• Against: Containment should be optional

Source Code format

• For: having a source code format such as XML or HTML allows for greater flexibility in creating a variety of display formats, with a greater likelihood of similarity between them
• Against: having the canonical format be in code ties us in to specific tools and/or tool support going forward