Comment Disposition for the FCD ballot of DSDL Part 4 (Amsterdam, 2005 May)

Some concepts used in the document are not clearly defined. There needs to be more explanatory matter to put the use of the described functions into context. Examples that deal with multiple instances of context would be helpful.

Accept. Modify example 2 in 5.2 to address this comment. Move the dcl of ns2 to to elements and also introduce another foo31 element having a different character content.

Note at end of list needs to be resolved and removed

Accept. Reference the latest RFCs for URIs and IRIs.

A cross reference to terms defined in ISO 8879 could usefully be added to the start of the list

Accept. List those terms borrowed from XML 1.0, the namespace recommendation, etc. at the beginning of Section 3.

Change "identifier" to "token"

Reject, since MIME RFCs use the term "identifier".

The term "action-mode pair" is not understandable, and seems to be self-referenceable.

Accept. Replace "action-mode pairs" with "actions".

The definition of slot node is circular. Not everyone understands what a slot is: the term should be defined in the standard.

Accept. The editor is instructed to change the definition of slot nodes. One possibility is to define slots as "slot: the position in a document from which an element or attribute section has been detached".

An introductory paragraph should be added to provide context for the following statements. This could be along the lines: "The following conventions have been adopted in the formal definitions provided in this standard:"

Accept, but replace "conventions" with "notations".

In definition of |e| change "number of elements" to "number of descendant elements".

Accept.

The term "place holder" is used consistently throughout the standard. The Oxford English Dictionary, however, gives the spelling of "placeholder" and does not offer an alternative with a space.

Accept.

The last phrase, "or that of some descendant of e" appears to be incorrect.

Accept. Replace "or that of some descendant element of e'" with "or occurs in the child sequence of some descendant element of e'".

This note is unclear.

Accept. Revise the note as follows: "An element e is not an ancestor of another element e' when the slot node for e' belongs to e."

In the last sentence the word "may" is used, implying that simplification is optional. Is the accurate, or should the word be replaced by "must" or, alternatively, "will"?

Reject. The simplification is indeed optional, since implementations are allowed to deviate from the simplification as long as the end result is identical.

The term "Qualified attributes" is not defined in clause 3, or in any of the XML documents referred to (which only ever refer to "unqualified attributes")

Accept. Replace "Qualified attributes" with "Attributes having namespace prefixes".

Change "ans" to "and"

Accept.

The term "Collision" in the heading is not used in the text. In the text the term "Competition" is used. Standardize on one term, e.g. "Competition within mode"

Accept. The new title for 6.4.12 is "Competition within mode".

It would appear that Case 3 can only exist in the case where w1=w2. The purpose of nsTail2 is, therefore, unclear.

Reject. ("asdf", "*") and ("asdf", "?") compete, although "*" is different from "?". However, the editor is instructed to provide examples of competition.

For Case 4 and Case 5 it should be made clear that the stripping of the first character occurs if it matches either w1 or w2.

Reject, since any character sequences matches w1 and w2. However, the editor is instructed to provide examples of competition.

In second paragraph "allow element" should be "reject element".

Accept.

Change "the child sequences ... is" to "the child sequences ... are"

Accept in principle, but replace "sequences" with "sequence".

See note above re "placeholder" and consider changing "placeHolder" to "placeholder".

Accept.

Change "denoted" to "known as".

Reject. "denoted" is very common in mathematical writing and is consistently used in this document.

The phrase "the parent element section" restricts the action to the immediately preceding element, rather than allowing it to affect all ancestor elements. Is this deliberate? If so, please explain why there is a (s) in the following "path(s)" as it would seem that only one path can be created.

Yes, this is liberate. path(s) is not plural, but rather a function call.

In the third Case 2 entry the phrase "some ending sequence" implies that any possible subsequence can match the specified sequence of Ncnames, not that all of the end must match it. Is this correct?

This "some" means "a certain".

In the fourth Case 2 what does the term "elder sibling" mean? If it means "preceding sibling" then change the term to the more standard terminology used throughout the XML specification.

Accept.

In the penultimate paragraph should "nsOrAny in mode" be "nsOrAny in mode'" (i.e. should mode be qualified by a quote?)

Reject. mode and mode' are different variables.

Change "is meant to choose" to "selects".

Accept.

Change "do not have" to "do not include".

Accept.

Change "an element" to "a foreign element".

Accept.

Change "esnXhtml2 and esnXhtml2" to "esnXhtml2 and esnXhtml3".

Accept.

Mention the new media type for RELAX NG compact syntax.

Accept.

wildcard rather than wildCard

Accept.