ISO/IEC JTC1/SC22/N1453 Subject: Comments arising from subgroup review of Section 7 To: WG5 From: Van Snyder Date: 2 August 2001 Four areas of concern arose from reviewing Section 7. Although subgroup has given some attention to each of these concerns, there has not been time to give them all of the attention that they may need. Therefore, the revisions proposed below may not actually be practical. 1. The treatment of the relational operators repeatedly mentions both their alphabetic and their symbolic spelling. WG5 prefers that the symbolic spelling given priority. WG5 suggests that this could be done at the lexical level in section 3, or that R713 [109:133-144] be re-written with six right-hand sides, each consisting of a single syntax term that specifies the semantic category of an operator symbol, e.g. R713 <> <> etc. with definitions for these syntax terms each in turn having two right-hand sides, e.g. R713a <> .EQ. <> == In either case, in the eight places where the relational operators appear in normative text, it is the case that both the alphabetic and the symbolic forms are specified. If the suggestion to modify R713 is taken up, in these places, the equivalent syntax term instead; otherwise the symbolic form should be used. In all but one of the places where the alphabetic form of a relational operator appears in an example, use the symbolic form instead. The one remaining example where alphabetic forms are used should illustrate that the alphabetic form is equivalent to the corresponding symbolic form. 2. The term "extension operation" is defined at [112:26-29]. The only place it is used is in table 7.7 [126:32,44]. The term "extension operator" is defined in the same paragraph, and never used anywhere. (The terms do not even appear in the glossary, even though they appear in bold-face type.) There is no advantage to (a) distinguishing extension operations from other defined operations, or (b) putting a "Category of operation" column in table 7.7. Delete all of [112:26-29], and the "Category of operation" column from table 7.7 [126:30-44]. Also see item 4 below. 3. Tables 7.5 and 7.6 are repetitive. Either one contains almost all of what one wants to know. Replace "7.5" by "7.6" at [125:6], delete [125:8-18], and insert the names of the operators below them in the heading (not the caption) of table 7.6 (which will become table 7.5) at [125:20+]: inclusive negation conjunction disjunction nonequivalence equivalence WG5 was divided equally (3-4-4) concerning the following observation: 4. Subclause 7.4 is entitled "Precedence of operators". It immediately launches into a discussion of precedence of _operations_. It is the operators upon which the precedence relations are defined, not the operations. The discussion should be only about precedence of _operators_. Furthermore, the entire subclause can be deduced from the syntax rules in subclause 7.1.1. As such, subclause 7.4, including table 7.7, should be a note. The relational operators could be specified either by reference to the syntax terms advocated in item 1, by specifying only their symbolic form, or this could be the only place where they are mentioned together.