PostLastCall

Actions tracking for Post Last Call.

RDF-WG possible impact points

1. Simple literals: simple literals are xsd:string in the abstract syntax
2. Plain literals with language tags
3. Graph identification
4. Turtle syntax

Query

Necessary before 2LC

• Comments
  • RC-5 // Draft response RC-5 - drafted - sent
• 2LC notes
• Check LC pub changes: msg

Not Andy:

• Check reformatting and edits in "17.4 Operator and Function Definitions"

Steve:

• "Order of evaluation of aggregates" 2011OctDec/0231
• Make sure it says functions and aggregates generating errors can be overridden (present in SPARQL 1.0 ; unclear in SPARQL 1.1 Query)

Completed:

• Prepare for RDF 1.1
  • REGEX, REPLACE to take all string-like forms as first argument (includes literals with language tag).
  • Check for others
  • Datatype(?x) to return rdf:langString on literals with a language tag
  • DATATYPE: Note change from SPARQL 1.0. (mark "at risk")
• Reformat function prototypes - reenable CSS
• Add STRAFTER, STRBEFORE, REPLACE
• Remove SHA224
• Update OWL reference (from comment IH-2)
• Function SUBSTR says "xsd:integer" : should say including derived types of integer.
• Grammar fix : ASK with GROUP BY, HAVING ASK now takes solution modifiers.
• Grammar notes need to have a point about aggregates and location.
• check CONSTRUCT WHERE short form.
• Sections 11 and 12: remove algebra. See CommentResponse:IH-1 (Steve)
• Grammar
  • Update with any escape sequence rules and any ORDER BY agg changes.
  • Consider ORDER BY count(*) (supported in SQL, just need to remove restriction in SPARQL)
    • Change grammar note 13 (which wasn't in 1LC at all)
    • "For each expression E in SELECT and each HAVING(E) in Q" ==> "For each expression E in SELECT and each HAVING(E) and each ORDER BY(E) in Q"
    • Put "Aggregate" in BuiltInCall (HAVING requires unnecessary brackets at the moment)
  • ConstructTemplate == TriplesTemplate
• "at-risk" text for escaping rules (red box)

• Reviews
  • ACTION-546 (Greg) Part 1, Part 2
  • ACTION-547 (Birte) Part 1 Part 2

Other editorial items

• Check scope rules and "BGP BIND"
• See SPARQL_Namespaces : Update page at the namespace URI of SPARQL functions
• IRIs for all functions (namespace document ?)

Desirable to do (may not happen ; time dependent)

• Move string literals argument type sections to near defn of simple literal.
• (unlikely) Query forms CONSTRUCT, ASK

Entailment

• Reviews
  • ACTION-552 (axel): Review Reply
  • Markus Kroetzsch: Review Reply

Update

• Use consistent presentation of GS in semantics 2011OctDec/0065 DONE
• Bug fixes on Definition in 4.2.5 msg - DONE
• What to do with the sentence suggested by comment DB-12, see also: msg DONE: sentence included.
• Description and defn of LOAD differ DONE: addressed with the rewording proposed in msg1, msg2. AP: I tend to interpret this change as non-LC critical, since the definitions haven't been changed, but only the inconsistent written text.
• Todo: remove @@@ at http://www.w3.org/2009/sparql/docs/update-1.1/Overview.xml#graphUpdate (editorial)
• LC pub changes: msg
• Link to http://www.w3.org/TR/rdf-sparql-query/#specifyingDataset needs to be changed to http://www.w3.org/TR/sparql11-query/#specifyingDataset also check for other links to the old spec! ( editorial)

• remove AT RISK note for COPY, ADD, MOVE

• Reviews
  • ACTION-554 (Matthew) DONE: Matt's review
  • ACTION-561 (olivier) DONE: Olivier's review

Service Description

• Edits in CommentResponse:DB-2
• Update "changes since last publication" section.
• Remove references to HTTP Protocol PUT operation.
• Update links to other SPARQL documents to point at /TR/ space.
• Add clarifying text for sd:propertyFeature (input from Axel and Chime?).
• Resolve Chime's comment re: "SPARQL Update 'subset'".
• Resolve Chime's comment re: ambiguity of "discovering" text.

• Reviews
  • ACTION-550 (Axel)
  • ACTION-551 (Chime)

SPARQL 1.1 Graph Store HTTP Protocol

• see http://lists.w3.org/Archives/Public/public-rdf-dawg/2011JulSep/0288.html

• Reviews
  • ACTION-563 (sandro)
  • ACTION-564 (andy)

JSON result format

• The JSON Note used "typed-literal"; the spec change/fixed that but should explicitly note it.

Protocol