W3C home > Mailing lists > Public > public-rif-wg@w3.org > May 2009

comments on RIF-Core

From: Stella Mitchell <stellamit@gmail.com>
Date: Mon, 11 May 2009 23:24:04 -0400
Message-ID: <d64b0f2c0905112024q579fe653mfa509cba8257cb99@mail.gmail.com>
To: RIF <public-rif-wg@w3.org>
Abstract:
-------------
     Also say that RIF-Core is specified as a
     specialization of RIF-PRD ?

1.0 Overview:
-------------------
    1st para:
          Says that Core is a language of production rules
          permitting only assert actions, but section 6.1 says
          that the symbol 'Assert' from PRD is excluded from Core.

    1st para, 1st sentence: (suggestion)
          This specification develops -->
          This specification describes

     4th para, 3rd sentence (suggestion)
           --> We therefore define a notion of safe RIF-Core
                rules, which is a subset of RIF-Core rules that
                can be executed using a forward chaining
                strategy, and we define conformance in terms of
                such safe rules.

     Example 1:
          update prefix directives with angle brackets around the iris

2.0
------
    In sections 2.1 through 2.5 it would be helpful if phrases like
    "alphabet of RIF-BLD presentation language," and
    "formulas of RIF-BLD" were links.


 2.5 Well-formed Formulas
-------------------------------------
       2nd para:
            Are these well-formedness tests supposed to be
            the well-formedness tests of RIF-BLD (as
            suggested by the paragraph above) just
            repeated in this document, or are they
            different Core well-formedness tests?

            item #2:
               Since the arity restriction was removed from
               BLD, shouldn't it also be removed from here
               (and the following paragraph)?

           BLD well-formedness has an additional condition
           related to external schemas, different from item #3,
           And also a condition about non rif:local constants
           in Document formulas with imports.

 2.6 EBNF Grammar for the Presentation Syntax of RIF-Core
--------------------------------------------------------------------------------------
      Currently the Mathematical English and the EBNF don't
      agree about whether membership is allowed in rule
      conclusions.

      GROUNDTERM production:
            'Expr' and some parens don't look like they're intended

      para following IRIMETA production:
             ANGLEBRACKIRI and CURIE -->
             ANGLEBRACKIRI, CURIE, CONSTSHORT and
             UNICODESTRING

 2.6.1
-------
      The reference to Example 2 in BLD needs to be updated.
      It's titled Example 2, but goes to Example 3, and
      both examples include non-Core syntax.

2.6.2
-------
     Change the text of the "Example 3" link to "Example 4",
     and the "Example 6" link to "Example 7"

     The linked examples need the prefix directives updated
     with angle brackets around the iri

 2.6.3
-------
     Change the text of the "Example 4" link to "Example 5"

     The linked example needs the prefix directives updated
     with angle brackets around the iri


 3.0 RIF-Core Semantics
-----------------------------------
      to the one --> to the semantics


 4.0 XML Serialization Syntax for RIF-Core
------------------------------------------------------------
      Change the title of the "Example 5" link to "Example 6",
      and the "Example 7" link to "Example 8," and update
      prefix directives with angle brackets in those examples

 5.1 Safeness
-------------------
     4th para, 1st bullet:
          pred:date-equal appears twice
            (same for the editor's note, and string-iri
            is replaced with string-equal there)

     para after the editor's note:
          "antecedents" isn't used anywhere else in the document.
          Change to "premises" to match the terminology in earlier
          sections?

     next para:
          outside of phi -->
          outside of phi in the rule?

    next para, 2nd bullet (disjunction):
          there is a set of phis, but only phi1 seems to come into
          play in the definition ?

 6.0
------
   In sections 6.1 through 6.6, it would be useful if phrases
   like "formulas of RIF-PRD," and "well-formedness test
   for RIF-PRD" were links.

 6.4 Annotations and Documents
-----------------------------------------------
     delete the last sentence because it is not true, and
     maybe the 2nd to last because all frame formulas in
     PRD are valid in RIF-Core?

 6.5 Well-formed Formulas
-------------------------------------
     The link to "well-formed formulas" in the last sentence
     doesn't seem relevant because goes to what may
     be the well-formedness tests for BLD that appear
     earlier in this document. Instead,  the
     "well-formedness test for RIF-PRD" in the
     previous sentence could be a link.

     Since BLD has a well-formedness condition
     about non rif:local constants in imported
     documents that PRD doesn't have, the
     defintion of core well-formedness in section
     2.5 may be a little different from the definition
     in this section.

   7.0 Conformance Clauses
--------------------------------------
     4th para:
         FLD and BLD use to specify a subset of L in the
         conformance clause for producers, but now they don't,
         but Core still mentions the subset.

     5th para:
        Says that the EBNF is informative and that the XML is
        normative, but doesn't say anything about the presentation
        syntax (as specified in the math english) -- BLD and FLD
        say it's normative.

     6th para:
          1st bullet:
               closed RIF condition -->
               closed RIF-Core condition

           2nd bullet:
               that do not match the syntax of Core ->
               that do not match the syntax of safe Core formulas ?


-Stella
Received on Tuesday, 12 May 2009 03:24:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:34:08 GMT