QT4 CG Meeting 048 Minutes 2023-10-03

Regrets NW.

• [X] Reece Dunn (RD)
• [X] Sasha Firsov (SF)
• [ ] Christian Grün (CG)
• [X] Joel Kalvesmaki (JK)
• [X] Michael Kay (MK)
• [X] John Lumley (JL) Scribe.
• [X] Dimitre Novatchev (DN)
• [X] Wendell Piez (WP)
• [X] Ed Porter (EP)
• [X] C. M. Sperberg-McQueen (MSM). Chair
• [ ] Norm Tovey-Walsh (NW).

• [ ] QT4CG-029-07: NW to open the next discussion of #397 with a demo from DN

See PR #449. Discussion planned for meeting 048, 3 October 2023.

• [ ] QT4CG-045-02: RD to address comments on HTML namespaces in another PR Still working on it
• [ ] QT4CG-046-01: MK to continue the work on #129 for the other specs (we accepted #703) Pending
• [ ] QT4CG-046-02: RD to draft the specification changes to allow record(*) Still working on it. MK thought he'd raised a PR on this one.
• [ ] QT4CG-046-03: MK to roll back the changes related to default namespaces for elments and types (issue #372) PR raised on this one - pending
• [ ] QT4CG-046-04: CG to flesh out changes related to annotations in other parts of the specs
• [ ] QT4CG-046-05: NW to updated parse-uri to use decode-from-uri (issue #566)
• [ ] QT4CG-047-01: MK to make a PR that renames break-when to split-when Pending

• MK: This is required more as a mode in 3.0+ introduces the distinct possibility that templates in a mode might be returning functions,maps, arrays, etc. This can be declared as a default for the templates within a mode,
• JL: It this a default or a constraint on consistency of the return type from anything in the mode.
• MK: Both
• DN: Should this be mandatory? Perhaps a bit restrictive - what about internal templates (not accessed from outside)?
• MSM: The subtyping rule for the @as. In some cases the mode might produce T1 and others T2. Use different modes, or a union type
• MK: Or don't use the feature. You might find yourself using more modes than normal.
• DN: I would find it too restrictive.
• MK: The driving use case was a schema-validator written in XSLT that use maps (records) VERY heavily. Every template rule in a mode had the same @as.
• JK: Note that the question I posed has now been resolved. Add a note suggesting feature designed for homogenous modes.
• JL: Very much in favour, especially with enclosed modes.
• WP: May be difficult if you're working with multiple mode templates.
• MSM: Homegeneity - useful if the mapping is of type T to type T, but perhaps less so for types X,Y,Z to type T, but with homogeneous cases it very much increases type checking.
• MK: Result type of a mode DOES invoked coercion, so e.g. @as="string" would invoke stringification (if permitted) or results.
• JL: Built-in actions (e.g. shallow-copy) - does it affect them.

ACTION QT4CG-048-01: MK - to identify what happens with the mode default rule behaviours.

• MK: It can also add compile-time checking of xsl:apply-templates.
• MSM: Add @as to apply-templates?
• MK: There is a proposal to add @as to xsl:sequence and others. Needs some more mulling other

Implicit general approval for letting this be status quo.

ACTION QT4CG-048-02: MK to clean up the proposal for adding @as to xsl:sequence and elsewhere

• MK: Given record type is accepted, but we need to look in some more detail. It is not just relevant to XSLT.
• RD: I don't think we've discussed record matching in XSMLT match patterns.
• JK: Record constructors in XSLT?
• MK: No proposals at present.
• MSM: Given not just XSLT - I would like to bring this back in a future meeting.
• MK: It should be in the XSLT spec section on patterns. In type patterns particular.

ACTION QT4CG-048-03: MK 5.4. - the section on type patterns - replace 3rd paragraph - belongs with predicate patterns. Fix cut-and-paste issues.

• MSM: Can we use ComplexItemType in these patterns
• MK: No - they are not an ItemType.

Some discussion on the WrappedItemType (type(….)) and short forms.

ACTION QT4CG-048-04: MK to use xs: prefix

Agreed it should be accepted for XSLT.