J3/08-264 To: J3 From: Malcolm Cohen Subject: Editorial changes that will be added to 019 soon. Date: 2008 August 12 1. Introduction This paper lists additional editorial changes the editor intends to add to the standing document 08-019, unless there is any objection. If there is a mistake in this list, please mention which items are possibly in error. Section 3 lists additional editorial changes that overlap 08-203. 2. Editorial Changes for 08-007r2 [3:1.5.4p3] "this part" -> "This part". {Capitalise first word of sentence.} [11:2.1.19-2.2.20] Move "corank" and "cosubscript" definitions to follow [13:2.1.29] "construct entity". {It wasn't in alphabetic order.} [20:2.1.95 "result variable" p1] Append "\ref{D12:Function subprogram}" i.e. "(12.6.2.2)" {This is the best reference for "result variable".} [61:4.5.2.4p2] After "Data entities" delete "in different scoping units". {We didn't intend to require data entities to be in different scoping units to activate this rule: sure it is the usual case, but when accessing the derived types via module renaming one can get them in the same scope.} [94:5.3.8.6] Index "named constant" here (here is "Implied-shape array"). {Implied-shape array is a significant thing for named constants.} [109:5.7.1.1, C583] After "dummy argument," insert "result variable name,", and delete "a function name, \obs{an entry name}, a result name,". {Simplification. An actual function name or entry name that is not a result variable name is already excluded by the syntax. Even when RESULT does not appear we still say that's the result variable, see first two sentences each of [303p4] and [306p3].} [same place] After "that has an allocatable" insert "or pointer", and delete "an object of a derived type that has a pointer at any level of component selection". {Simplification. Pointer components are already ultimate components, no matter what level of component selection is happening -- unless a higher level component was a pointer or allocatable, which is already excluded.} [same place, plus 110:C585] After "an automatic object," insert "a coarray," and delete C585 entirely. {Simplification. Coarray fits perfectly into the C583 list.} [112:5.7.2.1, C597] After "dummy argument," insert "result variable,", Delete "a function name, \obs{an entry name},", Before "a coarray" insert "or", Delete ", or a result name". {Same simplification as first edit to C583: what we mean here is the result variable, we don't have to say it thrice.} [124:C626] "nonprocedure pointer" -> "data pointer". {Use our defined term instead of negating its opposite.} [149:7.1.11p4(4)] After C_SIZEOF insert a space. {Typo.} [163:7.2.4.1p2] Near the end of the paragraph, replace "; it" by ". It". {Is the form "If A ... ; otherwise B ; C." really grammatical? I don't like it. Turn "; C." into a separate sentence to avoid confusion.} [175-181:8.1.7] Index "DO construct" throughout, not just the start. {The whole page range is relevant.} Unindex "construct!DO construct". {This is just a mistake. "construct!DO" is already indexed.} Unindex "DO CONCURRENT" - it occurs on exactly the same pages as "DO construct", indexing it separately achieves virtually nothing. {If DO CONCURRENT were indexed elsewhere, and DO were not indexed at that place, there would be a case for separate indexing.} [175:8.1.7.1p2] Replace "may be" by "can be", thrice. {Capability not permission.} [303:12.6.2.2] Index "result variable". Actually, index "result variable" throughout the document. {This term is completely unindexed apart from its definition.} [335:13.7.34p3] "integeror" -> "integer,". {Fix typo.} [387:17.7.155p3] After "ARRAY" change "a scalar or array" to "an array". {Fix typo applying edit at line 275 of 07-116r1.} ===END===