From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-0.8 required=5.0 tests=BAYES_00,INVALID_DATE,
	MSGID_SHORT autolearn=no autolearn_force=no version=3.4.4
Path: utzoo!mnetor!uunet!seismo!sundc!pitstop!sun!quintus!ok
From: ok@quintus.UUCP (Richard A. O'Keefe)
Newsgroups: comp.lang.ada
Subject: Re: Side effect
Message-ID: <863@cresswell.quintus.UUCP>
Date: 9 Apr 88 06:58:07 GMT
References: <14425@uflorida.cis.ufl.EDU>
Organization: Quintus Computer Systems, Mountain View, CA
List-Id: <comp.lang.ada>

In article <14425@uflorida.cis.ufl.EDU>, lius@beach.cis.ufl.edu (Scott Liu) writes:
>   ID: INTEGER := 0;
>   function NEXT return INTEGER is
>   begin
>     ID := ID+1;
>     return ID;
>   end;
> declare
>   type BIG is record X: INTEGER := NEXT; end record;  -- ???
>   A: array (1..5) of BIG;                             -- ???

{A is (1,2,3,4,5), not the expected (1,1,1,1,1).}

This is the way ADA is generally supposed to work.  For example,
	X, Y: INTEGER := NEXT;
is defined to be equivalent to
	X: INTEGER := NEXT;
	Y: INTEGER := NEXT;
In general, what matters is not how many copies of an initialiser you
write, but how many things are initialised:  an expression is evaluated
for _each_ variable it is used to initialise.  It's not the way I'd prefer
it, but at least it is consistent and predictable.

LRM 3.7 paras 5 and 11.
LRM 3.2.1 (note para 6).