From @sheila:swb@igor Fri Aug 26 20:33:11 1994 Return-Path: <@sheila:swb@igor> Received: from inmet.camb.inmet.com by dsd.camb.inmet.com (4.1/SMI-4.1) id AA26293; Fri, 26 Aug 94 20:33:11 EDT Received: from Rational.COM by inmet.camb.inmet.com (4.1/SMI-4.1) id AA18232; Fri, 26 Aug 94 20:33:07 EDT Received: from igor.Rational.COM by Rational.COM (5.0/rational-1.0.2) id AA11862; Fri, 26 Aug 1994 17:33:00 +0800 Received: from sheila by igor.Rational.COM (4.1/smh-1.0) id AA22779; Fri, 26 Aug 94 17:32:58 PDT Message-Id: <9408270032.AA22779@igor.Rational.COM> Date: Fri, 26 Aug 94 17:30:34 PDT From: swb@Rational.COM To: ada9x-mrt@inmet.com Cc: swb@igor Subject: withing private child units Hop-Count: 1 Content-Length: 1005 !topic withing private child units !reference AARM-10.1.2(8);5.0 !from Stephen Baird <> !discussion 10.1.2(8) states that "If a with_clause of a given compilation unit mentions a private child of some library unit, then the given comp_unit shall be either ... or the body ... of a ... descendant of that library unit". In this context, is a one-part (i.e. no forward declaration) library unit subprogram considered to be a body? How about a library unit subprogram rename which is the completion of a forward declaration? For example: package Parent is end Parent; private package Parent.Private_Child is end Parent.Private_Child; with Parent.Private_Child; -- legal? procedure Parent.One_Part_Subp_Child is begin null; end Parent.One_Part_Subp_Child; procedure Parent.Two_Part_Rename; procedure Foo; with Foo; with Parent.Private_Child; -- legal? procedure Parent.Two_Part_Rename renames Foo; Thank you.