From JANSHEP@torolab2.vnet.ibm.com Tue Mar 22 06:10:13 1994
Received: from vnet.ibm.com by dkuug.dk with SMTP id AA11610
  (5.65c8/IDA-1.4.4j for <sc22wg5@dkuug.dk>); Tue, 22 Mar 1994 17:11:41 +0100
Message-Id: <199403221611.AA11610@dkuug.dk>
Received: from TOROLAB2 by vnet.IBM.COM (IBM VM SMTP V2R2) with BSMTP id 3584;
   Tue, 22 Mar 94 11:08:44 EST
Date: Tue, 22 Mar 94 11:10:13 EST
From: "Janice Shepherd" <JANSHEP@torolab2.vnet.ibm.com>
To: sc22wg5@dkuug.dk
Subject: Corrigendum 2 ballot
X-Charset: ASCII
X-Char-Esc: 29

It has been pointed out to me that in the example below, I used
the wrong type of initialization statement. Sorry for the confusion.

>>|_|_| |x| 000097 Specification expression

>>Edits in a corrigendum and have to be applied manually by the reader.
>>Therefore wholesale rewrites of paragraphs should be avoided unless th
>>need is absolutely overwhelming. These should be defered to the 95
>>revision. Is anyone being seriously misled by the present wording?

>  The change in the wording makes clear that the following example
>is not valid. This would not be clear from the words as originally
>written in the standard.

    Subroutine X(I)
      integer aa(I)
!!    data ilocal /ubound(aa)/   ! not a valid data constant
      integer ilocal = ubound(aa)
      ...
    end subroutine

>The upper bound of 'aa' is not assumed -- 'aa' was not written as
>aa(:) nor aa(1:*), nor is 'aa' a pointer or an allocatable array.
>The next text clarifies that this example is not valid.

Janice C. Shepherd
IBM Canada, Toronto Lab.