ProposedElements/ModificationMarkup

Modification Markup

Investigating INS and DEL (and perhaps MOD)

1. what are the specific use cases for INS DEL and MOD

• to mark modified text
• to present deleted and inserted text in context
• to enable a user to parse that which has been deleted and that which has been inserted or modified

2. is a single element the best solution, or should there exist specific elements for insertion, deletion, and modification?

• pro: simplicity
• con: potential for abuse of attributes, unless strictly limited to the modification elements (INS, DEL, and MOD
  • Christophe Strobbe (K.U.Leuven - Research Group on Document Architectures) commented: I can't think of modifications that can't be handled with combinations of INS and DEL. Even changes in formatting can be handled with INS and DEL, even though a single element, MOD, may be more elegant.

3. why stick with just INS and DEL?

• Pro: This is the status quo and people are familiar with it.
  • INS and DEL are familiar to users who use change tracking in word processors like MS Word and OpenOffice.org Writer
• Con: may not satisfy all use cases; both block and inline elements (flow elements)

4. why elements rather than attributes?

• potential for abuse by authors attaching modification attributes to generic elements such as SPAN or DIV
  • Roland: might want both ways to do this; just add attribute to structural elements
• Roland's straw-man example: @diff, with values of add, chg, del

5. are there any unsatisfied use cases or requirements for modification markup?

• a mechanism to define that an insertion and a deletion belong together. For example, if you move a paragraph to another location in the document or if you replace a paragraph with other text in another location that is not obvious to the reader, it may be useful to be able to state that e.g. paragraph x (say, in chapter 2) has been replaced by paragraph y (say, in chapter 3).
  • there is a wai-xtech thread which illustrates the need for a means of binding insertions and deletions
    • it has been proposed that INS and DEL be bound by the for/id mechanism

6. where should INS and DEL be specified?

• XHTML2 text module

7. what is the nature of modification elements?

• currently INS and DEL are defined as "flow" elements, which can be used either inline or as a block-level element;

8. how granular should the modification elements be?

• is d<DEL>i</DEL><INS>o</INS>g legal?
  • answer: NO. The modification elements must span at least a word, and should not be used inside of a word to denote a change in spelling or case or tense;
• is <DEL id="blah1">dig</DEL> <INS for="blah1">dog</INS> best practice, or is d<DEL id="blah1">i</DEL><INS for="blah1">o</INS>g permissible, and is it any of our business?
• should <MOD src="http://www.foo.bar/document.html#p3s4">dog</MOD> be allowed? note that in this example, the global src attribute is used to point to the word which has been modified in the current version of the document; the src attribute is used to point directly to that which was modified;
  • use of MOD could be limited to use as an inline element
    • this might necessitate an attribute to state how modified, unless we use RDFa for this purpose;
      • Tina: MOD gets complicated - this is added this is deleted; MOD needs 2 sets of data -- what was, and what is; if can do with RDFa, fine, but don't think need that much complexity

---

Attributes Necessary for Modification Element(s)

1. a means of marking editorial changes;
2. a means of classifying an editorial change;
3. a means of conveying when and by whom the change was affected;
4. a means of binding an insertion with a deletion

Question 1: should the above-listed attributes be handled using [ RDFa], rather than element-specific attributes?

Question 2: what should be the mechanism used to add context and history to an INS or DEL element by using @src to link directly to the text that has been modified?

Question 3: would MOD with @src be handled differently than @src on other elements? should @href be used instead?

---