From helbig@man.ac.uk  Wed Mar 29 14:07:36 2000
Received: from mailhost.rug.nl (mailhost.rug.nl [129.125.4.6])
	by dkuug.dk (8.9.2/8.9.2) with ESMTP id OAA21144
	for <SC22WG5@dkuug.dk>; Wed, 29 Mar 2000 14:07:36 +0200 (CEST)
	(envelope-from helbig@man.ac.uk)
Received: from gladia.astro.rug.nl (gladia.astro.rug.nl [129.125.6.17])
	by mailhost.rug.nl (8.9.3/8.9.3) with SMTP id OAA26921
	for <SC22WG5@dkuug.dk>; Wed, 29 Mar 2000 14:07:35 +0200 (MET DST)
Date: Wed, 29 Mar 2000 14:06:57 +0100
Message-Id: <00032914065697@man.ac.uk>
From: helbig@man.ac.uk (Phillip Helbig)
To: J.S.Morgan@liverpool.ac.uk, SC22WG5@dkuug.dk
Subject: Re: (SC22WG5.1737) Interpretation 001
X-VMS-To: SMTP%"J.S.Morgan@liverpool.ac.uk"
X-VMS-Cc: SC22WG5@DKUUG.DK,HELBIG

> > I assume that they print the value 11 for j?
> 
> No, they all failed to compile!

:-(

> > > Module mod1
> > 
> > Compaq Fortran does this:
> 
> > j =            0

> > I assume this is the correct behaviour?
> 
> Yes, as long as j is initialised (by the compiler) to zero 
> but the important point, of course, is that j is not 
> exported from the module.

Right, it's probably initialised to 0.  However, had it been exported, 
the compiler would have claimed that it was declared twice:

$ type test.f90
Module mod1

INTEGER :: J

End Module mod1

Program prog1

USE mod1
INTEGER :: j

PRINT *, "j = ",j

End Program

$ fortran/stand=f95 test

INTEGER :: j
...........^
%F90-E-ERROR, This local name made accessible by a USE stmt appears in a specifi
cation stmt that respecifies the attributes.   [J]
at line number 10 in file DISK$SCRATCH:[HELBIG]TEST.F90;2
$

> > XLF90 also prints j = 0.