RIF Errata

Any change a W3C Recommendation must be approved by a chartered W3C Working Group and the W3C Advisory Committee. Until that happens, this "errata" page lists problems people have reported with the RIF documents, along with proposed solutions. To report a problem, send email to public-rif-comments@w3.org [archive]. We ask that RIF experts (especially the ones who produced the specifications) subscribe to that list and help review the proposed changes.

See the list archive for any unprocessed bug reports and in-progress discussion.

The RIF Second Edition addressed all the errata reported concerning the First Edition. For historical interest, see errata on the First Edition.

Proposed Erratum 1

Concerning: OWL 2 RL in RIF, 4.3.2 Property chain rule

Reinhard Schrage writes:

Below RIF Core rule set reads: 4.3.2 Property chain rule

 (* <#prp-spo2> *)
  Forall ?p ?last ?pc ?start (
    ?start[?p->?last] :- And (
        ?p[owl:propertyChainAxiom->?pc]
        _checkChain(?start ?pc ?last) ))

  Forall ?start ?pc ?last ?p ?tl (
    _checkChain(?start ?pc  ?last) :- And (
        ?pc[rdf:first->?p rdf:rest->?tl]
        ?start[?p->?next]
        _checkChain(?next ?tl ?last) ))

  Forall ?start ?pc ?last ?p (
    _checkChain(?start ?pc  ?last) :- And (
        ?pc[rdf:first->?p rdf:rest->rdf:nil]
        ?start[?p->?last] ))

Variable ?next is free in second universal rule, yet does not appear in Forall quantification part.

Am I missing something here, or shouldn't it read e.g.:

  Forall ?next ?start ?pc ?last ?p ?tl (
    _checkChain(?start ?pc  ?last) :- And (
        ?pc[rdf:first->?p rdf:rest->?tl]
        ?start[?p->?next]
        _checkChain(?next ?tl ?last) ))

Document editor Dave Reynolds replies:

Agreed, looks like a bug.