CVS difference for ais/ai-00117.txt

Differences between 1.13 and version 1.14
Log of other versions for file ais/ai-00117.txt

--- ais/ai-00117.txt	2000/08/01 05:39:33	1.13
+++ ais/ai-00117.txt	2001/03/22 22:24:53	1.14
@@ -396,9 +396,26 @@
 for a descendant of a generic formal untagged type.
 
 !ACATS test
+
 The response to question 1 is untestable, as the convention of such items
 rarely has a semantic effect.
-For the response to question 2:
+
+Response to question 2, part A & B:
+
+Create a B-Test and/or C-Test to check that convention of an inherited or
+overriding dispatching operation is the convention of the corresponding
+primitive operation of the type. Check that default convention of a new
+dispatching operation for a tagged type is the convention of the type.
+(Test, 6-0-2, ARG Letter Ballot, February 2001). Both tests would provide the
+best coverage. The obvious way to test this is to declare a package of
+convention C subprograms that are primitive on some type, then derive the
+type and check that the inherited ones can be used with a convention C access
+to subprogram type. Note that such a test will not be applicable to all
+implementations. Alternatively, a macro substitution of any supported language
+could be used instead of C, but the test still would not be applicable to
+every implementation. It is unclear whether Convention can be applied to a
+tagged type, so we must take care to allow proper Not Applicable behavior.
+
 A B-Test and C-Test should be constructed to check that the convention of
 dispatching routines is inherited. The obvious way to test this is to
 declare a package of convention C subprograms, and check that the inherited
@@ -406,10 +423,17 @@
 such a test will not be applicable to all implementations. Alternatively, a
 macro substitution of any supported language could be used instead of C, but
 the test still would not be applicable to every implementation.)
-For the response to question 3:
-The existing test C452001 has user-defined equality for tagged types. But it
-does not test dispatching of equality. A C-Test of dispatching equality of
-tagged types should be created (which would test this case as a side-effect).
+
+Response to question 2, part C:
+Create a C-Test to that operational and representation aspects of a partial
+view are the same as those of the full view. (No Test, 3-2-3 (2/3rd majority
+required, but not achieved), ARG Letter Ballot, February 2001). It is unlikely
+that a compiler would get this wrong (what over meaning could there be??)
+
+Response to question 3:
+Test C392013 checks that the "/=" implicitly declared with the
+declaration of "=" for a tagged type is legal and callable. (Test, 7-0-1,
+ARG Letter Ballot, February 2001).
 
 !appendix
 

Questions? Ask the ACAA Technical Agent