CVS difference for ai05s/ai05-0206-1.txt

Differences between 1.3 and version 1.4
Log of other versions for file ai05s/ai05-0206-1.txt

--- ai05s/ai05-0206-1.txt	2010/10/22 01:15:29	1.3
+++ ai05s/ai05-0206-1.txt	2011/03/17 06:16:41	1.4
@@ -1,4 +1,4 @@
-!standard  A(5/3)                                    10-10-21  AI05-0206-1/02
+!standard  A(5/3)                                    11-03-15  AI05-0206-1/03
 !standard  E.2(5)
 !standard  E.2(9)
 !standard  E.2(10)
@@ -120,6 +120,26 @@
 allowing packages to privately "with" preelaborated packages. There is no longer
 enough justification to keep the paragraph in the RM.
 
+We considered concerns about whether it was possible for a child package on a
+different partition to somehow "see" things in the private part of a Remote
+Types or RCI package. There are no issues here because;
+  a) For RCI packages, E.2.3(17) states that a child of an RCI library unit
+     has to belong to the same partition as the parent.
+  b) For Remote Type packages, E.2.2(7.a) explains that a separate copy of a
+     Remote Types package is included in each partition that references it, just
+     like a normal package. Like normal packages, any objects declared in the
+     private part, including RACW objects, are local entities that are only
+     accessible from the local partition. Such RACW objects may in fact denote
+     different remote objects on each partition, in the same way that ordinary
+     variables declared in a normal package can have different values in each
+     partition. E.2.2(7.a) goes on to state that the reason E.2.2(7) disallows
+     declarations of variables in the visible part of the library unit is only a
+     methodological restriction to avoid misleading by showing visible objects
+     in cases when no remote data access is actually being provided. Such
+     confusion is not likely to arise for variables declared in the private part
+     of a remote types parent package since they are private to the abstraction.
+
+
 --!corrigendum A(5/3)
 
 
@@ -880,5 +900,16 @@
 That's true of GNAT, though it is something we definitely intend to change,
 since it would be very useful to let a programmer know that pragma Pure or
 Preelaborate *could* be applied to a package.
+
+****************************************************************
+
+From: Brad Moore
+Sent: Tuesday, March 15, 2011  9:40 PM
+
+Attached is a new version of this AI. [This is version /03 of the AI. - Editor]
+
+All I changed was to add a paragraph to the end of the discussion
+section to answer the question that Steve Baird raised. I don't see any issues
+here.
 
 ****************************************************************

Questions? Ask the ACAA Technical Agent