This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
A small list of some (ir)relevant nitpicks: 1) In section 12, first Note, first para, it ends with: the format-numberFO30 function may be more suitable: see Section 4.7.2 fn:format-number FO30. which is essentially a double reference, the format-number function is already hyperlinked and then we point to it again. 2) 18.1.2 Streamability examples, we write: 4. The argument to count is a RelativePathExpr that takes the form of a motionless pattern[...] We don't have the "if an expression can be rewritten as a motionless pattern" rule anymore. We now have rules on Axis Steps and Path Expressions. It is still streamable, but for other reasons. 3) In section 21 Maps, we write: TODO: replace the error codes defined in this section by the error codes used in XPath 3.1. I think this has meanwhile been done (?), I don't see error codes other than the ones that apply to XSLT only. 4) In section Abstract, we write: TODO: describe the relationship to XPath 3.1 and related specs. I think this was done meanwhile, I see the relevant sections. 5) In section L Changes Since LCWD Under first item, also a "TODO": "TODO: add the streamability rules for new expressions and functions", but I think this one is still open (I scanned for fn:transform, but didn't find it). 6) In section B.1 Schema for the XML Representation of JSON, we write: <xs:simpleType name="numberType"> <xs:restriction base="xs:double"> <!-- exclude positive and negative infinity, and NaN --> <xs:minExclusive value="-INF"/> <xs:maxExclusive value="INF"/> </xs:restriction> </xs:simpleType> In the comment we say we exclude NaN, but we don't do so in the schema. Was this intended to be excluded? 7) In section 2.3.3 Apply-Templates Invocation, we write: The initial match selection. An API that chooses to maintain compatibility with previous versions of this specification should allow a singleton node to be provided, which is then used in two ways: the node itself acts as the initial match selection, and the document node of the containing document acts as the global context item. However, a singleton node does not need to be rooted at a document node. Should we be a bit more precise here (even though it is merely a suggestion to implementors)? 8) In section 24.1 Creating Final Result Trees, we write in the fourth bullet: The corresponding serialization parameter is not set (is "absent"). it is not possible [...] typo: it --> It
A small list of some (ir)relevant nitpicks: 1) In section 12, first Note, first para, it ends with: the format-numberFO30 function may be more suitable: see Section 4.7.2 fn:format-number FO30. * Fixed. Also in the immediately following para of the note, corrected "format-number" to "format-integer". 2) 18.1.2 Streamability examples, we write: 4. The argument to count is a RelativePathExpr that takes the form of a motionless pattern[...] We don't have the "if an expression can be rewritten as a motionless pattern" rule anymore. We now have rules on Axis Steps and Path Expressions. It is still streamable, but for other reasons. * I propose to replace (4). The argument to count is a RelativePathExpr that takes the form of a motionless pattern and appears in an inspection context (the argument to count()); the sweep of the RelativePathExpr and therefore of the entire body of the xsl:stream instruction is therefore consuming. with (4). The argument to count is a RelativePathExpr. Under the rules in 19.8.8.7 _Streamability of Path Expressions_, this expression is striding and consuming, which makes the call to count grounded and consuming. (5). Therefore the entire body of the xsl:stream instruction is grounded and consuming. 3) In section 21 Maps, we write: TODO: replace the error codes defined in this section by the error codes used in XPath 3.1. * Deleted the TODO. 4) In section Abstract, we write: TODO: describe the relationship to XPath 3.1 and related specs. I think this was done meanwhile, I see the relevant sections. * The Abstract itself still needs updating to reflect these changes. I've now made these changes. 5) In section L Changes Since LCWD. Under first item, also a "TODO": "TODO: add the streamability rules for new expressions and functions", but I think this one is still open (I scanned for fn:transform, but didn't find it). * This is covered adequately by bug #28937. 6) In section B.1 Schema for the XML Representation of JSON, we write: <xs:simpleType name="numberType"> <xs:restriction base="xs:double"> <!-- exclude positive and negative infinity, and NaN --> <xs:minExclusive value="-INF"/> <xs:maxExclusive value="INF"/> </xs:restriction> </xs:simpleType> In the comment we say we exclude NaN, but we don't do so in the schema. Was this intended to be excluded? * We include NaN implicitly. NaN < INF is false, so specifying a maxExclusive of INF automatically excludes NaN. 7) In section 2.3.3 Apply-Templates Invocation, we write: "The initial match selection. An API that chooses to maintain compatibility ... and the document node of the containing document acts as the global context item." However, a singleton node does not need to be rooted at a document node. Should we be a bit more precise here (even though it is merely a suggestion to implementors)? * Yes, should replace "the document node of the containing document" with "the root node of the containing tree". Fixed. 8) In section 24.1 Creating Final Result Trees, we write in the fourth bullet: The corresponding serialization parameter is not set (is "absent"). it is not possible [...] typo: it --> It * I think the reference is to 25.1 "Creating Secondary Results", and the relevant text has already been corrected in the current draft.
> * I think the reference is to 25.1 "Creating Secondary Results", and the > relevant text has already been corrected in the current draft. Indeed, 25.1 and indeed, it is corrected :).
The changes have been applied.